arm_compute v19.05
diff --git a/tests/validation/CL/ArithmeticAddition.cpp b/tests/validation/CL/ArithmeticAddition.cpp
index 04c9c85..bed04af 100644
--- a/tests/validation/CL/ArithmeticAddition.cpp
+++ b/tests/validation/CL/ArithmeticAddition.cpp
@@ -130,7 +130,7 @@
 
 TEST_SUITE(Quantized)
 TEST_SUITE(QASYMM8)
-DATA_TEST_CASE(Configuration, framework::DatasetMode::ALL, combine(datasets::SmallShapes(), framework::dataset::make("ConvertPolicy", { ConvertPolicy::SATURATE, ConvertPolicy::WRAP })),
+DATA_TEST_CASE(Configuration, framework::DatasetMode::ALL, combine(datasets::SmallShapes(), framework::dataset::make("ConvertPolicy", { ConvertPolicy::SATURATE })),
                shape, policy)
 {
     // Create tensors
@@ -155,7 +155,7 @@
 
 FIXTURE_DATA_TEST_CASE(RunSmall, CLArithmeticAdditionQuantizedFixture<uint8_t>, framework::DatasetMode::PRECOMMIT, combine(combine(combine(combine(combine(datasets::SmallShapes(),
                        ArithmeticAdditionQASYMM8Dataset),
-                       framework::dataset::make("ConvertPolicy", { ConvertPolicy::SATURATE, ConvertPolicy::WRAP })),
+                       framework::dataset::make("ConvertPolicy", { ConvertPolicy::SATURATE })),
                        framework::dataset::make("QuantizationInfo", { QuantizationInfo(5.f / 255.f, 20) })),
                        framework::dataset::make("QuantizationInfo", { QuantizationInfo(2.f / 255.f, 10) })),
                        framework::dataset::make("QuantizationInfo", { QuantizationInfo(1.f / 255.f, 5) }))
diff --git a/tests/validation/CL/ArithmeticSubtraction.cpp b/tests/validation/CL/ArithmeticSubtraction.cpp
index a8b54fd..796486b 100644
--- a/tests/validation/CL/ArithmeticSubtraction.cpp
+++ b/tests/validation/CL/ArithmeticSubtraction.cpp
@@ -131,7 +131,7 @@
 
 TEST_SUITE(Quantized)
 TEST_SUITE(QASYMM8)
-DATA_TEST_CASE(Configuration, framework::DatasetMode::ALL, combine(datasets::SmallShapes(), framework::dataset::make("ConvertPolicy", { ConvertPolicy::SATURATE, ConvertPolicy::WRAP })),
+DATA_TEST_CASE(Configuration, framework::DatasetMode::ALL, combine(datasets::SmallShapes(), framework::dataset::make("ConvertPolicy", { ConvertPolicy::SATURATE })),
                shape, policy)
 {
     // Create tensors
@@ -156,7 +156,7 @@
 
 FIXTURE_DATA_TEST_CASE(RunSmall, CLArithmeticSubtractionQuantizedFixture<uint8_t>, framework::DatasetMode::PRECOMMIT, combine(combine(combine(combine(combine(datasets::SmallShapes(),
                        ArithmeticSubtractionQASYMM8Dataset),
-                       framework::dataset::make("ConvertPolicy", { ConvertPolicy::SATURATE, ConvertPolicy::WRAP })),
+                       framework::dataset::make("ConvertPolicy", { ConvertPolicy::SATURATE })),
                        framework::dataset::make("QuantizationInfo", { QuantizationInfo(5.f / 255.f, 20) })),
                        framework::dataset::make("QuantizationInfo", { QuantizationInfo(2.f / 255.f, 10) })),
                        framework::dataset::make("QuantizationInfo", { QuantizationInfo(1.f / 255.f, 5) }))
diff --git a/tests/validation/CL/BatchToSpaceLayer.cpp b/tests/validation/CL/BatchToSpaceLayer.cpp
index e36663e..30b7817 100644
--- a/tests/validation/CL/BatchToSpaceLayer.cpp
+++ b/tests/validation/CL/BatchToSpaceLayer.cpp
@@ -52,21 +52,28 @@
 // clang-format off
 DATA_TEST_CASE(Validate, framework::DatasetMode::ALL, zip(zip(zip(
                framework::dataset::make("InputInfo", { TensorInfo(TensorShape(32U, 13U, 2U, 2U), 1, DataType::F32),
+                                                       TensorInfo(TensorShape(16U, 8U, 2U, 16U), 1, DataType::F32),    // blockx != blocky && blockx > blocky
+                                                       TensorInfo(TensorShape(16U, 8U, 2U, 16U), 1, DataType::F32),    // blockx != blocky && blocky > blockx
                                                        TensorInfo(TensorShape(32U, 13U, 2U, 2U), 1, DataType::F32),     // Mismatching data types
                                                        TensorInfo(TensorShape(32U, 13U, 2U, 2U), 1, DataType::F32),     // Wrong data type block shape
                                                        TensorInfo(TensorShape(32U, 13U, 2U, 2U, 4U), 1, DataType::F32), // Wrong tensor shape
                                                      }),
                framework::dataset::make("BlockShapeInfo",{ TensorInfo(TensorShape(2U, 2U), 1, DataType::S32),
+                                                       TensorInfo(TensorShape(2U, 4U), 1, DataType::S32),
+                                                      TensorInfo(TensorShape(4U, 2U), 1, DataType::S32),
                                                        TensorInfo(TensorShape(2U, 2U), 1, DataType::S32),
                                                        TensorInfo(TensorShape(2U, 2U), 1, DataType::F16),
                                                        TensorInfo(TensorShape(2U, 2U), 1, DataType::S32),
                                                      })),
                framework::dataset::make("OutputInfo",{ TensorInfo(TensorShape(32U, 13U, 2U, 2U), 1, DataType::F32),
+                                                       TensorInfo(TensorShape(64U, 16U, 2U, 1U), 1, DataType::F32),
+                                                       TensorInfo(TensorShape(32U, 32U, 2U, 1U), 1, DataType::F32),
+
                                                        TensorInfo(TensorShape(32U, 13U, 2U, 2U), 1, DataType::F16),
                                                        TensorInfo(TensorShape(32U, 13U, 2U, 2U), 1, DataType::F32),
                                                        TensorInfo(TensorShape(32U, 13U, 2U, 2U), 1, DataType::F32),
                                                      })),
-               framework::dataset::make("Expected", { true, false, false, false})),
+               framework::dataset::make("Expected", { true, true,true, false, false, false})),
                input_info, block_shape_info, output_info, expected)
 {
     bool has_error = bool(CLBatchToSpaceLayer::validate(&input_info.clone()->set_is_resizable(false), &block_shape_info.clone()->set_is_resizable(false), &output_info.clone()->set_is_resizable(false)));
@@ -74,18 +81,22 @@
 }
 DATA_TEST_CASE(ValidateStatic, framework::DatasetMode::ALL, zip(zip(zip(zip(
                framework::dataset::make("InputInfo", { TensorInfo(TensorShape(16U, 8U, 2U, 4U), 1, DataType::F32),
+                                                       TensorInfo(TensorShape(16U, 8U, 2U, 16U), 1, DataType::F32),    // blockx != blocky && blockx > blocky
+                                                       TensorInfo(TensorShape(16U, 8U, 2U, 16U), 1, DataType::F32),    // blockx != blocky && blocky > blockx
                                                        TensorInfo(TensorShape(16U, 8U, 2U, 4U), 1, DataType::F32),    // Mismatching data types
                                                        TensorInfo(TensorShape(16U, 8U, 2U, 4U), 1, DataType::F32),    // Negative block shapes
                                                        TensorInfo(TensorShape(32U, 16U, 2U, 4U, 4U), 1, DataType::F32), // Wrong tensor shape
                                                      }),
-               framework::dataset::make("BlockShapeX", { 2, 2, 2, 2 })),
-               framework::dataset::make("BlockShapeY", { 2, 2, -2, 2 })),
+               framework::dataset::make("BlockShapeX", { 2, 4, 2, 2, 2, 2 })),
+               framework::dataset::make("BlockShapeY", { 2, 2, 4, 2, -2, 2 })),
                framework::dataset::make("OutputInfo",{ TensorInfo(TensorShape(32U, 16U, 2U, 1U), 1, DataType::F32),
+                                                       TensorInfo(TensorShape(64U, 16U, 2U, 1U), 1, DataType::F32),
+                                                       TensorInfo(TensorShape(32U, 32U, 2U, 1U), 1, DataType::F32),
                                                        TensorInfo(TensorShape(32U, 16U, 2U, 1U), 1, DataType::F16),
                                                        TensorInfo(TensorShape(32U, 16U, 2U, 1U), 1, DataType::F32),
                                                        TensorInfo(TensorShape(32U, 8U, 2U, 1U), 1, DataType::F32),
                                                      })),
-               framework::dataset::make("Expected", { true, false, false, false})),
+               framework::dataset::make("Expected", { true, true,true, false, false, false})),
                input_info, block_shape_x, block_shape_y, output_info, expected)
 {
     bool has_error = bool(CLBatchToSpaceLayer::validate(&input_info.clone()->set_is_resizable(false), block_shape_x, block_shape_y, &output_info.clone()->set_is_resizable(false)));
diff --git a/tests/validation/CL/Convolution.cpp b/tests/validation/CL/Convolution.cpp
index 0c00ee3..9c33d45 100644
--- a/tests/validation/CL/Convolution.cpp
+++ b/tests/validation/CL/Convolution.cpp
@@ -54,14 +54,14 @@
     CLTensor dst = create_tensor<CLTensor>(shape, output_data_type);
 
     // Create conv matrix
-    int16_t conv[9];
+    std::array<int16_t, 9> conv = { 0 };
 
     ARM_COMPUTE_EXPECT(src.info()->is_resizable(), framework::LogLevel::ERRORS);
     ARM_COMPUTE_EXPECT(dst.info()->is_resizable(), framework::LogLevel::ERRORS);
 
     // Create and configure function
     CLConvolution3x3 convolution;
-    convolution.configure(&src, &dst, conv, 0, border_mode);
+    convolution.configure(&src, &dst, conv.data(), 0, border_mode);
 
     // Validate valid region
     const ValidRegion dst_valid_region = shape_to_valid_region(shape, (border_mode == BorderMode::UNDEFINED), BorderSize(filter_size / 2));
@@ -122,14 +122,14 @@
     CLTensor dst = create_tensor<CLTensor>(shape, output_data_type);
 
     // Create conv matrix
-    int16_t conv[25];
+    std::array<int16_t, 25> conv = { 0 };
 
     ARM_COMPUTE_EXPECT(src.info()->is_resizable(), framework::LogLevel::ERRORS);
     ARM_COMPUTE_EXPECT(dst.info()->is_resizable(), framework::LogLevel::ERRORS);
 
     // Create and configure function
     CLConvolution5x5 convolution;
-    convolution.configure(&src, &dst, conv, 0, border_mode);
+    convolution.configure(&src, &dst, conv.data(), 0, border_mode);
 
     // Validate valid region
     const ValidRegion dst_valid_region = shape_to_valid_region(shape, (border_mode == BorderMode::UNDEFINED), BorderSize(filter_size / 2));
@@ -190,14 +190,14 @@
     CLTensor dst = create_tensor<CLTensor>(shape, output_data_type);
 
     // Create conv matrix
-    int16_t conv[49];
+    std::array<int16_t, 49> conv = { 0 };
 
     ARM_COMPUTE_EXPECT(src.info()->is_resizable(), framework::LogLevel::ERRORS);
     ARM_COMPUTE_EXPECT(dst.info()->is_resizable(), framework::LogLevel::ERRORS);
 
     // Create and configure function
     CLConvolution7x7 convolution;
-    convolution.configure(&src, &dst, conv, 0, border_mode);
+    convolution.configure(&src, &dst, conv.data(), 0, border_mode);
 
     // Validate valid region
     const ValidRegion dst_valid_region = shape_to_valid_region(shape, (border_mode == BorderMode::UNDEFINED), BorderSize(filter_size / 2));
@@ -258,14 +258,14 @@
     CLTensor dst = create_tensor<CLTensor>(shape, output_data_type);
 
     // Create conv matrix
-    int16_t conv[81];
+    std::array<int16_t, 81> conv = { 0 };
 
     ARM_COMPUTE_EXPECT(src.info()->is_resizable(), framework::LogLevel::ERRORS);
     ARM_COMPUTE_EXPECT(dst.info()->is_resizable(), framework::LogLevel::ERRORS);
 
     // Create and configure function
     CLConvolution9x9 convolution;
-    convolution.configure(&src, &dst, conv, 0, border_mode);
+    convolution.configure(&src, &dst, conv.data(), 0, border_mode);
 
     // Validate valid region
     const ValidRegion dst_valid_region = shape_to_valid_region(shape, (border_mode == BorderMode::UNDEFINED), BorderSize(filter_size / 2));
@@ -328,14 +328,14 @@
     CLTensor dst = create_tensor<CLTensor>(shape, output_data_type);
 
     // Create conv matrix
-    int16_t conv[filter_width * filter_height];
+    std::vector<int16_t> conv(filter_width * filter_height);
 
     ARM_COMPUTE_EXPECT(src.info()->is_resizable(), framework::LogLevel::ERRORS);
     ARM_COMPUTE_EXPECT(dst.info()->is_resizable(), framework::LogLevel::ERRORS);
 
     // Create and configure function
     CLConvolutionRectangle convolution;
-    convolution.configure(&src, &dst, conv, filter_width, filter_height, 1, border_mode);
+    convolution.configure(&src, &dst, conv.data(), filter_width, filter_height, 1, border_mode);
 
     // Validate valid region
     const ValidRegion dst_valid_region = shape_to_valid_region(shape, (border_mode == BorderMode::UNDEFINED), BorderSize(filter_height / 2, filter_width / 2));
diff --git a/tests/validation/CL/ConvolutionLayer.cpp b/tests/validation/CL/ConvolutionLayer.cpp
index 41d2b7b..f1f9b59 100644
--- a/tests/validation/CL/ConvolutionLayer.cpp
+++ b/tests/validation/CL/ConvolutionLayer.cpp
@@ -46,7 +46,7 @@
 namespace
 {
 constexpr AbsoluteTolerance<float>  absolute_tolerance_float(0.0001f);    /**< Absolute Tolerance value for comparing reference's output against implementation's output for DataType::F32 */
-RelativeTolerance<float>            tolerance_f32(0.05f);                 /**< Tolerance value for comparing reference's output against implementation's output for DataType::F32 */
+RelativeTolerance<float>            tolerance_f32(0.1f);                  /**< Tolerance value for comparing reference's output against implementation's output for DataType::F32 */
 RelativeTolerance<half_float::half> tolerance_f16(half_float::half(0.2)); /**< Tolerance value for comparing reference's output against implementation's output for DataType::F16 */
 constexpr AbsoluteTolerance<float>  tolerance_qasymm8(1);                 /**< Tolerance value for comparing reference's output against implementation's output for quantized data types */
 constexpr float                     tolerance_num = 0.07f;                /**< Tolerance number */
diff --git a/tests/validation/CL/CropResize.cpp b/tests/validation/CL/CropResize.cpp
new file mode 100644
index 0000000..cacf405
--- /dev/null
+++ b/tests/validation/CL/CropResize.cpp
@@ -0,0 +1,184 @@
+/*
+ * Copyright (c) 2019 ARM Limited.
+ *
+ * SPDX-License-Identifier: MIT
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#include "arm_compute/core/Types.h"
+#include "arm_compute/runtime/CL/CLTensor.h"
+#include "arm_compute/runtime/CL/CLTensorAllocator.h"
+#include "arm_compute/runtime/CL/functions/CLCropResize.h"
+
+#include "tests/CL/CLAccessor.h"
+#include "tests/datasets/CropResizeDataset.h"
+#include "tests/framework/Asserts.h"
+#include "tests/framework/Macros.h"
+#include "tests/framework/datasets/Datasets.h"
+#include "tests/validation/Validation.h"
+#include "tests/validation/fixtures/CropResizeFixture.h"
+
+namespace arm_compute
+{
+namespace test
+{
+namespace validation
+{
+TEST_SUITE(CL)
+TEST_SUITE(CropResize)
+
+RelativeTolerance<float> tolerance_fp32(0.001f);
+
+template <typename T>
+using CLCropResizeFixture = CropResizeFixture<CLTensor, CLAccessor, CLCropResize, T>;
+
+// *INDENT-OFF*
+// clang-format off
+DATA_TEST_CASE(Validate, framework::DatasetMode::ALL, zip(zip(zip(zip(
+               framework::dataset::make("InputInfo", { TensorInfo(TensorShape(15U, 30U, 40U, 10U), 1, DataType::S32),
+                                                       TensorInfo(TensorShape(15U, 30U, 40U, 10U), 1, DataType::U8),  // Invalid input data type.
+                                                       TensorInfo(TensorShape(15U, 30U, 40U, 10U), 1, DataType::S32), // Invalid box_ind shape.
+                                                       TensorInfo(TensorShape(15U, 30U, 40U, 10U), 1, DataType::S32), // Invalid output shape.
+                                                       TensorInfo(TensorShape(15U, 30U, 40U, 10U), 1, DataType::S32), // Invalid output data type.
+                                                       TensorInfo(TensorShape(15U, 30U, 40U, 10U), 1, DataType::S32), // Invalid output shape.
+                                                       TensorInfo(TensorShape(15U, 30U, 40U, 10U), 1, DataType::S32), // Invalid boxes shape.
+                                                     }),
+               framework::dataset::make("BoxesInfo",{  TensorInfo(TensorShape(4, 20), 1, DataType::F32),
+                                                       TensorInfo(TensorShape(4, 20), 1, DataType::F32),
+                                                       TensorInfo(TensorShape(4, 20), 1, DataType::F32),
+                                                       TensorInfo(TensorShape(4, 20), 1, DataType::F32),
+                                                       TensorInfo(TensorShape(4, 20), 1, DataType::F32),
+                                                       TensorInfo(TensorShape(4, 20), 1, DataType::F32),
+                                                       TensorInfo(TensorShape(3, 20), 1, DataType::F32),
+                                                     })),
+               framework::dataset::make("BoxIndInfo",{ TensorInfo(TensorShape(20), 1, DataType::S32),
+                                                       TensorInfo(TensorShape(20), 1, DataType::S32),
+                                                       TensorInfo(TensorShape(10), 1, DataType::S32),
+                                                       TensorInfo(TensorShape(20), 1, DataType::S32),
+                                                       TensorInfo(TensorShape(20), 1, DataType::S32),
+                                                       TensorInfo(TensorShape(20), 1, DataType::S32),
+                                                       TensorInfo(TensorShape(20), 1, DataType::S32),
+                                                     })),
+               framework::dataset::make("OutputInfo",{ TensorInfo(TensorShape(15U, 5, 5, 20U), 1, DataType::F32),
+                                                       TensorInfo(TensorShape(15U, 5, 5, 20U), 1, DataType::F32),
+                                                       TensorInfo(TensorShape(15U, 5, 5, 20U), 1, DataType::F32),
+                                                       TensorInfo(TensorShape(15U, 5, 5, 10U), 1, DataType::F32),
+                                                       TensorInfo(TensorShape(15U, 5, 5, 20U), 1, DataType::S32),
+                                                       TensorInfo(TensorShape(5U, 5, 5, 20U), 1, DataType::F32),
+                                                       TensorInfo(TensorShape(15U, 5, 5, 20U), 1, DataType::F32),
+                                                     })),
+               framework::dataset::make("Expected", { true, false, false, false, false, false, false})),
+               input, boxes, box_ind, output, expected)
+{
+    ARM_COMPUTE_EXPECT(bool(CLCropResize::validate(&input.clone()->set_data_layout(DataLayout::NHWC).set_is_resizable(false),
+                                                   &boxes.clone()->set_is_resizable(false),
+                                                   &box_ind.clone()->set_is_resizable(false),
+                                                   &output.clone()->set_data_layout(DataLayout::NHWC).set_is_resizable(false),
+                                                   Coordinates2D{ 5, 5 }, InterpolationPolicy::BILINEAR, 100)) == expected, framework::LogLevel::ERRORS);
+}
+// clang-format on
+// *INDENT-ON*
+
+TEST_SUITE(Float)
+#ifdef __ARM_FEATURE_FP16_VECTOR_ARITHMETIC
+TEST_SUITE(F16)
+FIXTURE_DATA_TEST_CASE(RunSmall,
+                       CLCropResizeFixture<half>,
+                       framework::DatasetMode::PRECOMMIT,
+                       combine(datasets::SmallCropResizeDataset(),
+                               combine(framework::dataset::make("IsOutOfBounds", { true, false }),
+                                       framework::dataset::make("DataType", DataType::F16))))
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference, tolerance_fp32, 0.01);
+}
+TEST_SUITE_END() // F16
+#endif           /* __ARM_FEATURE_FP16_VECTOR_ARITHMETIC */
+
+TEST_SUITE(F32)
+FIXTURE_DATA_TEST_CASE(RunSmall,
+                       CLCropResizeFixture<float>,
+                       framework::DatasetMode::PRECOMMIT,
+                       combine(datasets::SmallCropResizeDataset(),
+                               combine(framework::dataset::make("IsOutOfBounds", { true, false }),
+                                       framework::dataset::make("DataType", DataType::F32))))
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference, tolerance_fp32, 0.01);
+}
+TEST_SUITE_END() // F32
+TEST_SUITE_END() // Float
+
+TEST_SUITE(U16)
+FIXTURE_DATA_TEST_CASE(RunSmall,
+                       CLCropResizeFixture<uint16_t>,
+                       framework::DatasetMode::PRECOMMIT,
+                       combine(datasets::SmallCropResizeDataset(),
+                               combine(framework::dataset::make("IsOutOfBounds", { true, false }),
+                                       framework::dataset::make("DataType", DataType::U16))))
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference, tolerance_fp32, 0.01);
+}
+TEST_SUITE_END() // U16
+
+TEST_SUITE(S16)
+FIXTURE_DATA_TEST_CASE(RunSmall,
+                       CLCropResizeFixture<int16_t>,
+                       framework::DatasetMode::PRECOMMIT,
+                       combine(datasets::SmallCropResizeDataset(),
+                               combine(framework::dataset::make("IsOutOfBounds", { true, false }),
+                                       framework::dataset::make("DataType", DataType::S16))))
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference, tolerance_fp32, 0.01);
+}
+TEST_SUITE_END() // S16
+
+TEST_SUITE(U32)
+FIXTURE_DATA_TEST_CASE(RunSmall,
+                       CLCropResizeFixture<uint32_t>,
+                       framework::DatasetMode::PRECOMMIT,
+                       combine(datasets::SmallCropResizeDataset(),
+                               combine(framework::dataset::make("IsOutOfBounds", { true, false }),
+                                       framework::dataset::make("DataType", DataType::U32))))
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference, tolerance_fp32, 0.01);
+}
+TEST_SUITE_END() // U32
+
+TEST_SUITE(S32)
+FIXTURE_DATA_TEST_CASE(RunSmall,
+                       CLCropResizeFixture<int32_t>,
+                       framework::DatasetMode::PRECOMMIT,
+                       combine(datasets::SmallCropResizeDataset(),
+                               combine(framework::dataset::make("IsOutOfBounds", { true, false }),
+                                       framework::dataset::make("DataType", DataType::S32))))
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference, tolerance_fp32, 0.01);
+}
+TEST_SUITE_END() // S32
+
+TEST_SUITE_END() // CropResize
+TEST_SUITE_END() // CL
+} // namespace validation
+} // namespace test
+} // namespace arm_compute
diff --git a/tests/validation/CL/DeconvolutionLayer.cpp b/tests/validation/CL/DeconvolutionLayer.cpp
index 31852c8..db93beb 100644
--- a/tests/validation/CL/DeconvolutionLayer.cpp
+++ b/tests/validation/CL/DeconvolutionLayer.cpp
@@ -58,10 +58,13 @@
 const auto data3x3_precommit = datasets::SmallDeconvolutionShapes() * framework::dataset::make("StrideX", 1, 2) * framework::dataset::make("StrideY", 1, 2) * framework::dataset::make("PadX", 0, 2)
                                * framework::dataset::make("PadY", 0, 2) * framework::dataset::make("NumKernels", { 3 });
 
+const auto data2x2_precommit = datasets::SmallDeconvolutionShapes() * framework::dataset::make("StrideX", 2) * framework::dataset::make("StrideY", 2) * framework::dataset::make("PadX", 1)
+                               * framework::dataset::make("PadY", 1) * framework::dataset::make("NumKernels", { 3 });
+
 const auto data1x1 = datasets::SmallDeconvolutionShapes() * framework::dataset::make("StrideX", 1, 4) * framework::dataset::make("StrideY", 1, 4) * framework::dataset::make("PadX", 0, 1)
                      * framework::dataset::make("PadY", 0, 1) * framework::dataset::make("NumKernels", { 3 });
 
-const auto data_layouts_dataset = framework::dataset::make("DataLayout", { DataLayout::NCHW });
+const auto data_layouts_dataset = framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC });
 } // namespace
 
 TEST_SUITE(CL)
@@ -119,7 +122,7 @@
                                                 1U,
                                                 0U,
                                             })),
-    framework::dataset::make("Expected", { false, false, false, false, false, true })),
+    framework::dataset::make("Expected", { false, false, false, false, true, true })),
     input_info, weights_info, bias_info, output_info, pad_info, ax, ay, expected)
 {
     bool is_valid = bool(CLDeconvolutionLayer::validate(&input_info.clone()->set_is_resizable(false), &weights_info.clone()->set_is_resizable(false), &bias_info.clone()->set_is_resizable(false), &output_info.clone()->set_is_resizable(false), pad_info, ax, ay));
@@ -135,6 +138,9 @@
 using CLDeconvolutionLayerFixture3x3 = DeconvolutionValidationFixture<CLTensor, CLAccessor, CLDeconvolutionLayer, T, 3, 3>;
 
 template <typename T>
+using CLDeconvolutionLayerFixture2x2 = DeconvolutionValidationFixture<CLTensor, CLAccessor, CLDeconvolutionLayer, T, 2, 2>;
+
+template <typename T>
 using CLDeconvolutionLayerFixture1x1 = DeconvolutionValidationFixture<CLTensor, CLAccessor, CLDeconvolutionLayer, T, 1, 1>;
 
 TEST_SUITE(Float)
@@ -164,6 +170,15 @@
 }
 TEST_SUITE_END() // W3x3
 
+TEST_SUITE(W2x2)
+FIXTURE_DATA_TEST_CASE(RunSmall, CLDeconvolutionLayerFixture2x2<float>, framework::DatasetMode::PRECOMMIT, combine(combine(data2x2_precommit, framework::dataset::make("DataType", DataType::F32)),
+                                                                                                                   data_layouts_dataset))
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference, tolerance_fp32);
+}
+TEST_SUITE_END() // W2x2
+
 TEST_SUITE(W1x1)
 FIXTURE_DATA_TEST_CASE(Run, CLDeconvolutionLayerFixture1x1<float>, framework::DatasetMode::NIGHTLY, combine(combine(data1x1, framework::dataset::make("DataType", DataType::F32)),
                                                                                                             data_layouts_dataset))
@@ -200,6 +215,15 @@
 }
 TEST_SUITE_END() // W3x3
 
+TEST_SUITE(W2x2)
+FIXTURE_DATA_TEST_CASE(RunSmall, CLDeconvolutionLayerFixture2x2<half>, framework::DatasetMode::PRECOMMIT, combine(combine(data2x2_precommit, framework::dataset::make("DataType", DataType::F16)),
+                                                                                                                  data_layouts_dataset))
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference, tolerance_f16);
+}
+TEST_SUITE_END() // W2x2
+
 TEST_SUITE(W1x1)
 FIXTURE_DATA_TEST_CASE(Run, CLDeconvolutionLayerFixture1x1<half>, framework::DatasetMode::NIGHTLY, combine(combine(data1x1, framework::dataset::make("DataType", DataType::F16)), data_layouts_dataset))
 {
@@ -218,6 +242,9 @@
 using CLDeconvolutionLayerQuantizedFixture3x3 = DeconvolutionValidationQuantizedFixture<CLTensor, CLAccessor, CLDeconvolutionLayer, T, 3, 3>;
 
 template <typename T>
+using CLDeconvolutionLayerQuantizedFixture2x2 = DeconvolutionValidationQuantizedFixture<CLTensor, CLAccessor, CLDeconvolutionLayer, T, 2, 2>;
+
+template <typename T>
 using CLDeconvolutionLayerQuantizedFixture1x1 = DeconvolutionValidationQuantizedFixture<CLTensor, CLAccessor, CLDeconvolutionLayer, T, 1, 1>;
 
 TEST_SUITE(Quantized)
@@ -253,6 +280,17 @@
 }
 TEST_SUITE_END() // W3x3
 
+TEST_SUITE(W2x2)
+FIXTURE_DATA_TEST_CASE(RunSmall, CLDeconvolutionLayerQuantizedFixture2x2<uint8_t>, framework::DatasetMode::PRECOMMIT, combine(combine(combine(data2x2_precommit, framework::dataset::make("DataType",
+                       DataType::QASYMM8)),
+                       data_layouts_dataset),
+                       framework::dataset::make("QuantizationInfo", QuantizationInfo(2.f / 255.f, 0))))
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference, tolerance_fp32);
+}
+TEST_SUITE_END() // W2x2
+
 TEST_SUITE(W1x1)
 FIXTURE_DATA_TEST_CASE(Run, CLDeconvolutionLayerQuantizedFixture1x1<uint8_t>, framework::DatasetMode::NIGHTLY, combine(combine(combine(data1x1, framework::dataset::make("DataType",
                                                                                                                        DataType::QASYMM8)),
diff --git a/tests/validation/CL/DepthConcatenateLayer.cpp b/tests/validation/CL/DepthConcatenateLayer.cpp
index 01477f9..8cbfda3 100644
--- a/tests/validation/CL/DepthConcatenateLayer.cpp
+++ b/tests/validation/CL/DepthConcatenateLayer.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2017-2018 ARM Limited.
+ * Copyright (c) 2017-2019 ARM Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -24,14 +24,14 @@
 #include "arm_compute/core/Types.h"
 #include "arm_compute/runtime/CL/CLTensor.h"
 #include "arm_compute/runtime/CL/CLTensorAllocator.h"
-#include "arm_compute/runtime/CL/functions/CLDepthConcatenateLayer.h"
+#include "arm_compute/runtime/CL/functions/CLConcatenateLayer.h"
 #include "tests/CL/CLAccessor.h"
 #include "tests/datasets/ShapeDatasets.h"
 #include "tests/framework/Asserts.h"
 #include "tests/framework/Macros.h"
 #include "tests/framework/datasets/Datasets.h"
 #include "tests/validation/Validation.h"
-#include "tests/validation/fixtures/DepthConcatenateLayerFixture.h"
+#include "tests/validation/fixtures/ConcatenateLayerFixture.h"
 
 namespace arm_compute
 {
@@ -68,13 +68,13 @@
     inputs_vector_info.emplace_back(std::move(input_info2));
 
     std::vector<ITensorInfo *> inputs_vector_info_raw;
+    inputs_vector_info_raw.reserve(inputs_vector_info.size());
     for(auto &input : inputs_vector_info)
     {
         inputs_vector_info_raw.emplace_back(&input);
     }
 
-    bool is_valid = bool(CLDepthConcatenateLayer::validate(inputs_vector_info_raw,
-                                                           &output_info.clone()->set_is_resizable(false)));
+    bool is_valid = bool(CLConcatenateLayer::validate(inputs_vector_info_raw, &output_info.clone()->set_is_resizable(false), 2));
     ARM_COMPUTE_EXPECT(is_valid == expected, framework::LogLevel::ERRORS);
 }
 // clang-format on
@@ -83,37 +83,38 @@
 TEST_CASE(Configuration, framework::DatasetMode::ALL)
 {
     // Create tensors
-    CLTensor src1 = create_tensor<CLTensor>(TensorShape(32U, 32U, 128U), DataType::F32, 1);
-    CLTensor src2 = create_tensor<CLTensor>(TensorShape(32U, 32U, 32U), DataType::F32, 1);
+    CLTensor src1 = create_tensor<CLTensor>(TensorShape(128U, 32U, 32U), DataType::F32, 1);
+    CLTensor src2 = create_tensor<CLTensor>(TensorShape(128U, 32U, 32U), DataType::F32, 1);
+    CLTensor src3 = create_tensor<CLTensor>(TensorShape(128U, 32U, 32U), DataType::F32, 1);
     CLTensor dst;
 
     ARM_COMPUTE_EXPECT(src1.info()->is_resizable(), framework::LogLevel::ERRORS);
     ARM_COMPUTE_EXPECT(src2.info()->is_resizable(), framework::LogLevel::ERRORS);
+    ARM_COMPUTE_EXPECT(src3.info()->is_resizable(), framework::LogLevel::ERRORS);
     ARM_COMPUTE_EXPECT(dst.info()->is_resizable(), framework::LogLevel::ERRORS);
 
     // Create and configure function
-    CLDepthConcatenateLayer concat_layer;
+    CLConcatenateLayer concat_layer;
 
-    concat_layer.configure({ &src1, &src2 }, &dst);
-
-    // Validate valid region
-    const ValidRegion valid_region = shape_to_valid_region(TensorShape(32U, 32U, 160U));
-    validate(dst.info()->valid_region(), valid_region);
+    concat_layer.configure({ &src1, &src2, &src3 }, &dst, 2);
 }
-
 template <typename T>
-using CLDepthConcatenateLayerFixture = DepthConcatenateLayerValidationFixture<CLTensor, ICLTensor, CLAccessor, CLDepthConcatenateLayer, T>;
+using CLDepthConcatenateLayerFixture = ConcatenateLayerValidationFixture<CLTensor, ICLTensor, CLAccessor, CLConcatenateLayer, T>;
 
 TEST_SUITE(Float)
 TEST_SUITE(FP16)
-FIXTURE_DATA_TEST_CASE(RunSmall, CLDepthConcatenateLayerFixture<half>, framework::DatasetMode::PRECOMMIT, combine(datasets::Small2DShapes(), framework::dataset::make("DataType",
-                                                                                                                  DataType::F16)))
+FIXTURE_DATA_TEST_CASE(RunSmall, CLDepthConcatenateLayerFixture<half>, framework::DatasetMode::PRECOMMIT, combine(combine(concat(datasets::Small3DShapes(), datasets::Tiny4DShapes()),
+                                                                                                                  framework::dataset::make("DataType",
+                                                                                                                          DataType::F16)),
+                                                                                                                  framework::dataset::make("Axis", 2)))
 {
     // Validate output
     validate(CLAccessor(_target), _reference);
 }
-FIXTURE_DATA_TEST_CASE(RunLarge, CLDepthConcatenateLayerFixture<half>, framework::DatasetMode::NIGHTLY, combine(datasets::Large2DShapes(), framework::dataset::make("DataType",
-                                                                                                                DataType::F16)))
+FIXTURE_DATA_TEST_CASE(RunLarge, CLDepthConcatenateLayerFixture<half>, framework::DatasetMode::NIGHTLY, combine(combine(concat(datasets::Large3DShapes(), datasets::Small4DShapes()),
+                                                                                                                        framework::dataset::make("DataType",
+                                                                                                                                DataType::F16)),
+                                                                                                                framework::dataset::make("Axis", 2)))
 {
     // Validate output
     validate(CLAccessor(_target), _reference);
@@ -121,14 +122,17 @@
 TEST_SUITE_END()
 
 TEST_SUITE(FP32)
-FIXTURE_DATA_TEST_CASE(RunSmall, CLDepthConcatenateLayerFixture<float>, framework::DatasetMode::PRECOMMIT, combine(datasets::Small2DShapes(), framework::dataset::make("DataType",
-                                                                                                                   DataType::F32)))
+FIXTURE_DATA_TEST_CASE(RunSmall, CLDepthConcatenateLayerFixture<float>, framework::DatasetMode::PRECOMMIT, combine(combine(concat(datasets::Small3DShapes(), datasets::Tiny4DShapes()),
+                                                                                                                   framework::dataset::make("DataType",
+                                                                                                                           DataType::F32)),
+                                                                                                                   framework::dataset::make("Axis", 2)))
 {
     // Validate output
     validate(CLAccessor(_target), _reference);
 }
-FIXTURE_DATA_TEST_CASE(RunLarge, CLDepthConcatenateLayerFixture<float>, framework::DatasetMode::NIGHTLY, combine(datasets::DepthConcatenateLayerShapes(), framework::dataset::make("DataType",
-                                                                                                                 DataType::F32)))
+FIXTURE_DATA_TEST_CASE(RunLarge, CLDepthConcatenateLayerFixture<float>, framework::DatasetMode::NIGHTLY, combine(combine(datasets::ConcatenateLayerShapes(), framework::dataset::make("DataType",
+                                                                                                                 DataType::F32)),
+                                                                                                                 framework::dataset::make("Axis", 2)))
 {
     // Validate output
     validate(CLAccessor(_target), _reference);
@@ -136,17 +140,19 @@
 TEST_SUITE_END()
 TEST_SUITE_END()
 
-
 TEST_SUITE(Quantized)
 TEST_SUITE(QASYMM8)
-FIXTURE_DATA_TEST_CASE(RunSmall, CLDepthConcatenateLayerFixture<uint8_t>, framework::DatasetMode::PRECOMMIT, combine(datasets::Small2DShapes(), framework::dataset::make("DataType",
-                                                                                                                     DataType::QASYMM8)))
+FIXTURE_DATA_TEST_CASE(RunSmall, CLDepthConcatenateLayerFixture<uint8_t>, framework::DatasetMode::PRECOMMIT, combine(combine(concat(datasets::Small3DShapes(), datasets::Tiny4DShapes()),
+                                                                                                                     framework::dataset::make("DataType",
+                                                                                                                             DataType::QASYMM8)),
+                                                                                                                     framework::dataset::make("Axis", 2)))
 {
     // Validate output
     validate(CLAccessor(_target), _reference);
 }
-FIXTURE_DATA_TEST_CASE(RunLarge, CLDepthConcatenateLayerFixture<uint8_t>, framework::DatasetMode::NIGHTLY, combine(datasets::DepthConcatenateLayerShapes(), framework::dataset::make("DataType",
-                                                                                                                   DataType::QASYMM8)))
+FIXTURE_DATA_TEST_CASE(RunLarge, CLDepthConcatenateLayerFixture<uint8_t>, framework::DatasetMode::NIGHTLY, combine(combine(datasets::ConcatenateLayerShapes(), framework::dataset::make("DataType",
+                                                                                                                   DataType::QASYMM8)),
+                                                                                                                   framework::dataset::make("Axis", 2)))
 {
     // Validate output
     validate(CLAccessor(_target), _reference);
@@ -154,7 +160,6 @@
 TEST_SUITE_END()
 TEST_SUITE_END()
 
-
 TEST_SUITE_END()
 TEST_SUITE_END()
 } // namespace validation
diff --git a/tests/validation/CL/DepthwiseConvolutionLayer.cpp b/tests/validation/CL/DepthwiseConvolutionLayer.cpp
index dd2d9f3..274a0f5 100644
--- a/tests/validation/CL/DepthwiseConvolutionLayer.cpp
+++ b/tests/validation/CL/DepthwiseConvolutionLayer.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2017-2018 ARM Limited.
+ * Copyright (c) 2017-2019 ARM Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -18,7 +18,7 @@
  * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
  * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
  * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
- * OUT OF OR IN CONCLCTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
  * SOFTWARE.
  */
 #include "arm_compute/core/Types.h"
@@ -28,6 +28,7 @@
 #include "tests/CL/CLAccessor.h"
 #include "tests/PaddingCalculator.h"
 #include "tests/datasets/DepthwiseConvolutionLayerDataset.h"
+#include "tests/datasets/DilatedDepthwiseConvolutionLayerDataset.h"
 #include "tests/framework/Asserts.h"
 #include "tests/framework/Macros.h"
 #include "tests/framework/datasets/Datasets.h"
@@ -48,6 +49,13 @@
 constexpr float                      tolerance_num = 0.05f;                 /**< Tolerance number */
 
 const auto depth_multipliers = framework::dataset::make("DepthMultiplier", { 1, 2, 3 });
+
+//Activation Functions
+const auto ActivationFunctionsDataset = framework::dataset::make("ActivationInfo",
+{
+    ActivationLayerInfo(),
+    ActivationLayerInfo(ActivationLayerInfo::ActivationFunction::RELU)
+});
 } // namespace
 
 TEST_SUITE(CL)
@@ -55,7 +63,7 @@
 
 // *INDENT-OFF*
 // clang-format off
-DATA_TEST_CASE(Validate3x3, framework::DatasetMode::ALL, zip(zip(zip(zip(zip(zip(zip(
+DATA_TEST_CASE(Validate3x3, framework::DatasetMode::ALL, zip(zip(zip(zip(zip(zip(zip(zip(
                framework::dataset::make("InputInfo", { TensorInfo(TensorShape(32U, 18U, 2U), 1, DataType::F32),     // Mismatching data type input/weights
                                                        TensorInfo(TensorShape(32U, 18U, 3U), 1, DataType::F32),     // Mismatching input feature maps
                                                        TensorInfo(TensorShape(32U, 18U, 2U), 1, DataType::F32),     // Unsupported weights dimensions
@@ -66,6 +74,8 @@
                                                        TensorInfo(TensorShape(32U, 18U, 2U), 1, DataType::F32),     // Invalid biases dimensions
                                                        TensorInfo(TensorShape(32U, 18U, 2U), 1, DataType::F32),     // Invalid output size
                                                        TensorInfo(TensorShape(27U, 13U, 2U), 1, DataType::F32),     // Window shrink
+                                                       TensorInfo(TensorShape(32U, 18U, 8U), 1, DataType::F32),     // patch size bigger than input width
+                                                       TensorInfo(TensorShape(32U, 18U, 8U), 1, DataType::F32),     // dilation < 1
                                                        TensorInfo(TensorShape(32U, 18U, 8U), 1, DataType::F32),
                                                        TensorInfo(TensorShape(50U, 32U, 8U), 1, DataType::QASYMM8),
                                                      }),
@@ -80,6 +90,8 @@
                                                          TensorInfo(TensorShape(3U, 3U, 2U), 1, DataType::F32),
                                                          TensorInfo(TensorShape(3U, 3U, 2U), 1, DataType::F32),
                                                          TensorInfo(TensorShape(3U, 3U, 16U), 1, DataType::F32),
+                                                         TensorInfo(TensorShape(3U, 3U, 16U), 1, DataType::F32),
+                                                         TensorInfo(TensorShape(3U, 3U, 16U), 1, DataType::F32),
                                                          TensorInfo(TensorShape(3U, 3U, 24U), 1, DataType::QASYMM8),
                                                        })),
                framework::dataset::make("BiasesInfo", { TensorInfo(TensorShape(2U), 1, DataType::F32),
@@ -93,6 +105,8 @@
                                                         TensorInfo(TensorShape(2U), 1, DataType::F32),
                                                         TensorInfo(TensorShape(2U), 1, DataType::F32),
                                                         TensorInfo(TensorShape(16U), 1, DataType::F32),
+                                                        TensorInfo(TensorShape(16U), 1, DataType::F32),
+                                                        TensorInfo(TensorShape(16U), 1, DataType::F32),
                                                         TensorInfo(TensorShape(24U), 1, DataType::S32),
                                                       })),
                framework::dataset::make("OutputInfo", { TensorInfo(TensorShape(30U, 16U, 2U), 1, DataType::F32),
@@ -106,6 +120,8 @@
                                                         TensorInfo(TensorShape(32U, 18U, 2U), 1, DataType::F32),
                                                         TensorInfo(TensorShape(25U, 11U, 2U), 1, DataType::F32),
                                                         TensorInfo(TensorShape(30U, 16U, 16U), 1, DataType::F32),
+                                                        TensorInfo(TensorShape(30U, 16U, 16U), 1, DataType::F32),
+                                                        TensorInfo(TensorShape(30U, 16U, 16U), 1, DataType::F32),
                                                         TensorInfo(TensorShape(48U, 30U, 24U), 1, DataType::QASYMM8),
                                                       })),
                framework::dataset::make("ConvInfo", { PadStrideInfo(1, 1, 0, 0),
@@ -120,6 +136,8 @@
                                                       PadStrideInfo(1, 1, 0, 0),
                                                       PadStrideInfo(1, 1, 0, 0),
                                                       PadStrideInfo(1, 1, 0, 0),
+                                                      PadStrideInfo(1, 1, 0, 0),
+                                                      PadStrideInfo(1, 1, 0, 0),
                                                      })),
                framework::dataset::make("DepthMultiplier", { 1,
                                                              1,
@@ -132,6 +150,8 @@
                                                              1,
                                                              1,
                                                              2,
+                                                             2,
+                                                             2,
                                                              3,
                                                             })),
                 framework::dataset::make("ActivationInfo", { ActivationLayerInfo(),
@@ -145,22 +165,41 @@
                                                              ActivationLayerInfo(),
                                                              ActivationLayerInfo(),
                                                              ActivationLayerInfo(),
+                                                             ActivationLayerInfo(),
+                                                             ActivationLayerInfo(),
                                                              ActivationLayerInfo(ActivationLayerInfo::ActivationFunction::RELU),
                                                            })),
-               framework::dataset::make("Expected", { false, false, false, false, false, false, false, false, false, false, true, true })),
-               input_info, weights_info, biases_info, output_info, conv_info, depth_multiplier, act_info, expected)
+                framework::dataset::make("Dilation", { Size2D(1U, 1U),
+                                                       Size2D(1U, 1U),
+                                                       Size2D(1U, 1U),
+                                                       Size2D(1U, 1U),
+                                                       Size2D(1U, 1U),
+                                                       Size2D(1U, 1U),
+                                                       Size2D(1U, 1U),
+                                                       Size2D(1U, 1U),
+                                                       Size2D(1U, 1U),
+                                                       Size2D(1U, 1U),
+                                                       Size2D(20U, 1U),
+                                                       Size2D(0U, 1U),
+                                                       Size2D(1U, 1U),
+                                                       Size2D(1U, 1U),
+                                                       })),
+               framework::dataset::make("Expected", { false, false, false, false, false, false, false, false, false, false, false, false, true, true })),
+               input_info, weights_info, biases_info, output_info, conv_info, depth_multiplier, act_info, dilation, expected)
 {
-    bool is_valid = bool(CLDepthwiseConvolutionLayer3x3::validate(&input_info.clone()->set_is_resizable(false), &weights_info.clone()->set_is_resizable(false), &biases_info.clone()->set_is_resizable(false), &output_info.clone()->set_is_resizable(false), conv_info, depth_multiplier, act_info));
+    bool is_valid = bool(CLDepthwiseConvolutionLayer3x3::validate(&input_info.clone()->set_is_resizable(false), &weights_info.clone()->set_is_resizable(false), &biases_info.clone()->set_is_resizable(false), &output_info.clone()->set_is_resizable(false), conv_info, depth_multiplier, act_info,GPUTarget::MIDGARD, dilation));
     ARM_COMPUTE_EXPECT(is_valid == expected, framework::LogLevel::ERRORS);
 }
 
-DATA_TEST_CASE(ValidateGeneric, framework::DatasetMode::ALL, zip(zip(zip(zip(zip(zip(
+DATA_TEST_CASE(ValidateGeneric, framework::DatasetMode::ALL, zip(zip(zip(zip(zip(zip(zip(
                 framework::dataset::make("InputInfo", { TensorInfo(TensorShape(27U, 13U, 2U), 1, DataType::F32),    // Mismatching data type input/weights
                                                         TensorInfo(TensorShape(27U, 13U, 3U), 1, DataType::F32),    // Mismatching input feature maps
                                                         TensorInfo(TensorShape(27U, 13U, 2U), 1, DataType::F32),    // Mismatching depth multiplier
                                                         TensorInfo(TensorShape(27U, 13U, 2U), 1, DataType::F32),    // Invalid biases size
                                                         TensorInfo(TensorShape(27U, 13U, 2U), 1, DataType::F32),    // Invalid biases dimensions
                                                         TensorInfo(TensorShape(27U, 13U, 2U), 1, DataType::F32),    // Invalid output size
+                                                        TensorInfo(TensorShape(27U, 13U, 8U), 1, DataType::F32),    // patch size bigger than input width
+                                                        TensorInfo(TensorShape(27U, 13U, 8U), 1, DataType::F32),    // dilation < 1
                                                         TensorInfo(TensorShape(27U, 13U, 8U), 1, DataType::F32),
                                                         TensorInfo(TensorShape(32U, 13U, 8U), 1, DataType::QASYMM8),
                                                       }),
@@ -171,6 +210,8 @@
                                                           TensorInfo(TensorShape(3U, 3U, 2U), 1, DataType::F32),
                                                           TensorInfo(TensorShape(3U, 3U, 2U), 1, DataType::F32),
                                                           TensorInfo(TensorShape(3U, 3U, 16U), 1, DataType::F32),
+                                                          TensorInfo(TensorShape(3U, 3U, 16U), 1, DataType::F32),
+                                                          TensorInfo(TensorShape(3U, 3U, 16U), 1, DataType::F32),
                                                           TensorInfo(TensorShape(3U, 3U, 24U), 1, DataType::QASYMM8),
                                                         })),
                 framework::dataset::make("BiasesInfo", { TensorInfo(TensorShape(2U), 1, DataType::F32),
@@ -180,6 +221,8 @@
                                                          TensorInfo(TensorShape(2U, 2U), 1, DataType::F32),
                                                          TensorInfo(TensorShape(2U), 1, DataType::F32),
                                                          TensorInfo(TensorShape(16U), 1, DataType::F32),
+                                                         TensorInfo(TensorShape(16U), 1, DataType::F32),
+                                                         TensorInfo(TensorShape(16U), 1, DataType::F32),
                                                          TensorInfo(TensorShape(24U), 1, DataType::S32),
                                                        })),
                 framework::dataset::make("OutputInfo", { TensorInfo(TensorShape(25U, 11U, 2U), 1, DataType::F32),
@@ -189,6 +232,8 @@
                                                          TensorInfo(TensorShape(25U, 11U, 2U), 1, DataType::F32),
                                                          TensorInfo(TensorShape(27U, 13U, 2U), 1, DataType::F32),
                                                          TensorInfo(TensorShape(25U, 11U, 16U), 1, DataType::F32),
+                                                         TensorInfo(TensorShape(25U, 11U, 16U), 1, DataType::F32),
+                                                         TensorInfo(TensorShape(25U, 11U, 16U), 1, DataType::F32),
                                                          TensorInfo(TensorShape(32U, 11U, 24U), 1, DataType::QASYMM8),
                                                        })),
                 framework::dataset::make("ConvInfo", { PadStrideInfo(1, 1, 0, 0),
@@ -198,6 +243,8 @@
                                                        PadStrideInfo(1, 1, 0, 0),
                                                        PadStrideInfo(1, 1, 0, 0),
                                                        PadStrideInfo(1, 1, 0, 0),
+                                                       PadStrideInfo(1, 1, 0, 0),
+                                                       PadStrideInfo(1, 1, 0, 0),
                                                        PadStrideInfo(1, 1, 1, 0),
                                                       })),
                 framework::dataset::make("DepthMultiplier", { 1,
@@ -207,12 +254,25 @@
                                                               1,
                                                               1,
                                                               2,
+                                                              2,
+                                                              2,
                                                               3,
                                                              })),
-                framework::dataset::make("Expected", { false, false, false, false, false, false, true, true })),
-                input_info, weights_info, biases_info, output_info, conv_info, depth_multiplier, expected)
+                framework::dataset::make("Dilation", { Size2D(1U, 1U),
+                                                              Size2D(1U, 1U),
+                                                              Size2D(1U, 1U),
+                                                              Size2D(1U, 1U),
+                                                              Size2D(1U, 1U),
+                                                              Size2D(1U, 1U),
+                                                              Size2D(20U, 1U),
+                                                              Size2D(0U, 1U),
+                                                              Size2D(1U, 1U),
+                                                              Size2D(1U, 1U),
+                                                             })),
+                framework::dataset::make("Expected", { false, false, false, false, false, false, false, false, true, true })),
+                input_info, weights_info, biases_info, output_info, conv_info, depth_multiplier, dilation, expected)
 {
-    bool is_valid = bool(CLDepthwiseConvolutionLayer::validate(&input_info.clone()->set_is_resizable(false), &weights_info.clone()->set_is_resizable(false), &biases_info.clone()->set_is_resizable(false), &output_info.clone()->set_is_resizable(false), conv_info, depth_multiplier));
+    bool is_valid = bool(CLDepthwiseConvolutionLayer::validate(&input_info.clone()->set_is_resizable(false), &weights_info.clone()->set_is_resizable(false), &biases_info.clone()->set_is_resizable(false), &output_info.clone()->set_is_resizable(false), conv_info, depth_multiplier,ActivationLayerInfo(), dilation));
     ARM_COMPUTE_EXPECT(is_valid == expected, framework::LogLevel::ERRORS);
 }
 // clang-format on
@@ -226,127 +286,269 @@
 TEST_SUITE(W3x3)
 TEST_SUITE(NCHW)
 FIXTURE_DATA_TEST_CASE(RunSmall, CLDepthwiseConvolutionLayerFixture<half>, framework::DatasetMode::ALL,
-                       combine(combine(combine(framework::dataset::concat(datasets::SmallDepthwiseConvolutionLayerDataset3x3(),
-                                                                          datasets::SmallDepthwiseConvolutionLayerDataset3x3NCHW()),
-                                               depth_multipliers),
-                                       framework::dataset::make("DataType",
-                                                                DataType::F16)),
-                               framework::dataset::make("DataLayout", DataLayout::NCHW)))
+                       combine(combine(combine(combine(framework::dataset::concat(datasets::SmallDepthwiseConvolutionLayerDataset3x3(),
+                                                                                  datasets::SmallDepthwiseConvolutionLayerDataset3x3NCHW()),
+                                                       depth_multipliers),
+                                               framework::dataset::make("DataType",
+                                                                        DataType::F16)),
+                                       framework::dataset::make("DataLayout", DataLayout::NCHW)),
+                               ActivationFunctionsDataset))
 {
     validate(CLAccessor(_target), _reference, tolerance_f16);
 }
-FIXTURE_DATA_TEST_CASE(RunLarge, CLDepthwiseConvolutionLayerFixture<half>, framework::DatasetMode::NIGHTLY, combine(combine(combine(datasets::LargeDepthwiseConvolutionLayerDataset3x3(),
+FIXTURE_DATA_TEST_CASE(RunLarge, CLDepthwiseConvolutionLayerFixture<half>, framework::DatasetMode::NIGHTLY, combine(combine(combine(combine(datasets::LargeDepthwiseConvolutionLayerDataset3x3(),
                                                                                                                     depth_multipliers),
                                                                                                                     framework::dataset::make("DataType",
                                                                                                                             DataType::F16)),
-                                                                                                                    framework::dataset::make("DataLayout", DataLayout::NCHW)))
+                                                                                                                    framework::dataset::make("DataLayout", DataLayout::NCHW)),
+                                                                                                                    ActivationFunctionsDataset))
 {
     validate(CLAccessor(_target), _reference, tolerance_f16);
 }
-TEST_SUITE_END()
+TEST_SUITE(Dilation)
+FIXTURE_DATA_TEST_CASE(RunSmall, CLDepthwiseConvolutionLayerFixture<half>, framework::DatasetMode::ALL, combine(combine(combine(combine(datasets::SmallDepthwiseDilatedConvolutionLayerDataset3x3(),
+                                                                                                                        depth_multipliers),
+                                                                                                                        framework::dataset::make("DataType",
+                                                                                                                                DataType::F16)),
+                                                                                                                        framework::dataset::make("DataLayout", { DataLayout::NCHW })),
+                                                                                                                ActivationFunctionsDataset))
+{
+    validate(CLAccessor(_target), _reference, tolerance_f16);
+}
+FIXTURE_DATA_TEST_CASE(RunLarge, CLDepthwiseConvolutionLayerFixture<half>, framework::DatasetMode::NIGHTLY, combine(combine(combine(combine(datasets::LargeDepthwiseDilatedConvolutionLayerDataset3x3(),
+                                                                                                                    depth_multipliers),
+                                                                                                                    framework::dataset::make("DataType",
+                                                                                                                            DataType::F16)),
+                                                                                                                    framework::dataset::make("DataLayout", { DataLayout::NCHW })),
+                                                                                                                    ActivationFunctionsDataset))
+{
+    validate(CLAccessor(_target), _reference, tolerance_f16);
+}
+TEST_SUITE_END() // Dilation
+TEST_SUITE_END() // NCHW
 
 TEST_SUITE(NHWC)
 FIXTURE_DATA_TEST_CASE(RunSmall, CLDepthwiseConvolutionLayerFixture<half>, framework::DatasetMode::ALL,
-                       combine(combine(combine(datasets::SmallDepthwiseConvolutionLayerDataset3x3(),
-                                               depth_multipliers),
-                                       framework::dataset::make("DataType",
-                                                                DataType::F16)),
-                               framework::dataset::make("DataLayout", DataLayout::NHWC)))
+                       combine(combine(combine(combine(datasets::SmallDepthwiseConvolutionLayerDataset3x3(),
+                                                       depth_multipliers),
+                                               framework::dataset::make("DataType",
+                                                                        DataType::F16)),
+                                       framework::dataset::make("DataLayout", DataLayout::NHWC)),
+                               ActivationFunctionsDataset))
 {
     validate(CLAccessor(_target), _reference, tolerance_f16);
 }
-FIXTURE_DATA_TEST_CASE(RunLarge, CLDepthwiseConvolutionLayerFixture<half>, framework::DatasetMode::NIGHTLY, combine(combine(combine(datasets::LargeDepthwiseConvolutionLayerDataset3x3(),
+FIXTURE_DATA_TEST_CASE(RunLarge, CLDepthwiseConvolutionLayerFixture<half>, framework::DatasetMode::NIGHTLY, combine(combine(combine(combine(datasets::LargeDepthwiseConvolutionLayerDataset3x3(),
                                                                                                                     depth_multipliers),
                                                                                                                     framework::dataset::make("DataType",
                                                                                                                             DataType::F16)),
-                                                                                                                    framework::dataset::make("DataLayout", DataLayout::NHWC)))
+                                                                                                                    framework::dataset::make("DataLayout", DataLayout::NHWC)),
+                                                                                                                    ActivationFunctionsDataset))
 {
     validate(CLAccessor(_target), _reference, tolerance_f16);
 }
-TEST_SUITE_END()
-TEST_SUITE_END()
-
-TEST_SUITE(Generic)
-FIXTURE_DATA_TEST_CASE(RunSmall, CLDepthwiseConvolutionLayerFixture<half>, framework::DatasetMode::ALL, combine(combine(combine(datasets::SmallDepthwiseConvolutionLayerDataset(), depth_multipliers),
+TEST_SUITE(Dilation)
+FIXTURE_DATA_TEST_CASE(RunSmall, CLDepthwiseConvolutionLayerFixture<half>, framework::DatasetMode::ALL, combine(combine(combine(combine(datasets::SmallDepthwiseDilatedConvolutionLayerDataset3x3(),
+                                                                                                                        depth_multipliers),
                                                                                                                         framework::dataset::make("DataType",
                                                                                                                                 DataType::F16)),
-                                                                                                                framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })))
+                                                                                                                        framework::dataset::make("DataLayout", { DataLayout::NHWC })),
+                                                                                                                ActivationFunctionsDataset))
 {
-    validate(CLAccessor(_target), _reference, tolerance_f16, tolerance_num);
+    validate(CLAccessor(_target), _reference, tolerance_f16);
 }
-FIXTURE_DATA_TEST_CASE(RunLarge, CLDepthwiseConvolutionLayerFixture<half>, framework::DatasetMode::NIGHTLY, combine(combine(combine(datasets::LargeDepthwiseConvolutionLayerDataset(),
+FIXTURE_DATA_TEST_CASE(RunLarge, CLDepthwiseConvolutionLayerFixture<half>, framework::DatasetMode::NIGHTLY, combine(combine(combine(combine(datasets::LargeDepthwiseDilatedConvolutionLayerDataset3x3(),
                                                                                                                     depth_multipliers),
                                                                                                                     framework::dataset::make("DataType",
                                                                                                                             DataType::F16)),
-                                                                                                                    framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })))
+                                                                                                                    framework::dataset::make("DataLayout", { DataLayout::NHWC })),
+                                                                                                                    ActivationFunctionsDataset))
+{
+    validate(CLAccessor(_target), _reference, tolerance_f16);
+}
+TEST_SUITE_END() // Dilation
+TEST_SUITE_END() // NHWC
+TEST_SUITE_END() // W3x3
+
+TEST_SUITE(Generic)
+FIXTURE_DATA_TEST_CASE(RunSmall, CLDepthwiseConvolutionLayerFixture<half>, framework::DatasetMode::ALL, combine(combine(combine(combine(datasets::SmallDepthwiseConvolutionLayerDataset(),
+                                                                                                                        depth_multipliers),
+                                                                                                                        framework::dataset::make("DataType",
+                                                                                                                                DataType::F16)),
+                                                                                                                        framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })),
+                                                                                                                ActivationFunctionsDataset))
 {
     validate(CLAccessor(_target), _reference, tolerance_f16, tolerance_num);
 }
-TEST_SUITE_END()
-TEST_SUITE_END()
+FIXTURE_DATA_TEST_CASE(RunLarge, CLDepthwiseConvolutionLayerFixture<half>, framework::DatasetMode::NIGHTLY, combine(combine(combine(combine(datasets::LargeDepthwiseConvolutionLayerDataset(),
+                                                                                                                    depth_multipliers),
+                                                                                                                    framework::dataset::make("DataType",
+                                                                                                                            DataType::F16)),
+                                                                                                                    framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })),
+                                                                                                                    ActivationFunctionsDataset))
+{
+    validate(CLAccessor(_target), _reference, tolerance_f16, tolerance_num);
+}
+
+TEST_SUITE(Dilation)
+FIXTURE_DATA_TEST_CASE(RunSmall, CLDepthwiseConvolutionLayerFixture<half>, framework::DatasetMode::ALL, combine(combine(combine(combine(datasets::SmallDepthwiseDilatedConvolutionLayerDataset(),
+                                                                                                                        depth_multipliers),
+                                                                                                                        framework::dataset::make("DataType",
+                                                                                                                                DataType::F16)),
+                                                                                                                        framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })),
+                                                                                                                ActivationFunctionsDataset))
+{
+    validate(CLAccessor(_target), _reference, tolerance_f16, tolerance_num);
+}
+FIXTURE_DATA_TEST_CASE(RunLarge, CLDepthwiseConvolutionLayerFixture<half>, framework::DatasetMode::NIGHTLY, combine(combine(combine(combine(datasets::LargeDepthwiseDilatedConvolutionLayerDataset(),
+                                                                                                                    depth_multipliers),
+                                                                                                                    framework::dataset::make("DataType",
+                                                                                                                            DataType::F16)),
+                                                                                                                    framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })),
+                                                                                                                    ActivationFunctionsDataset))
+{
+    validate(CLAccessor(_target), _reference, tolerance_f16, tolerance_num);
+}
+TEST_SUITE_END() // Dilation
+TEST_SUITE_END() // Generic
+TEST_SUITE_END() // FP16
 
 TEST_SUITE(FP32)
 TEST_SUITE(W3x3)
 TEST_SUITE(NCHW)
 FIXTURE_DATA_TEST_CASE(RunSmall, CLDepthwiseConvolutionLayerFixture<float>, framework::DatasetMode::ALL,
-                       combine(combine(combine(framework::dataset::concat(datasets::SmallDepthwiseConvolutionLayerDataset3x3(),
-                                                                          datasets::SmallDepthwiseConvolutionLayerDataset3x3NCHW()),
-                                               depth_multipliers),
-                                       framework::dataset::make("DataType",
-                                                                DataType::F32)),
-                               framework::dataset::make("DataLayout", DataLayout::NCHW)))
+                       combine(combine(combine(combine(framework::dataset::concat(datasets::SmallDepthwiseConvolutionLayerDataset3x3(),
+                                                                                  datasets::SmallDepthwiseConvolutionLayerDataset3x3NCHW()),
+                                                       depth_multipliers),
+                                               framework::dataset::make("DataType",
+                                                                        DataType::F32)),
+                                       framework::dataset::make("DataLayout", DataLayout::NCHW)),
+                               ActivationFunctionsDataset))
 {
     validate(CLAccessor(_target), _reference, tolerance_f32);
 }
-FIXTURE_DATA_TEST_CASE(RunLarge, CLDepthwiseConvolutionLayerFixture<float>, framework::DatasetMode::NIGHTLY, combine(combine(combine(datasets::LargeDepthwiseConvolutionLayerDataset3x3(),
+FIXTURE_DATA_TEST_CASE(RunLarge, CLDepthwiseConvolutionLayerFixture<float>, framework::DatasetMode::NIGHTLY, combine(combine(combine(combine(datasets::LargeDepthwiseConvolutionLayerDataset3x3(),
                                                                                                                      depth_multipliers),
                                                                                                                      framework::dataset::make("DataType",
                                                                                                                              DataType::F32)),
-                                                                                                                     framework::dataset::make("DataLayout", DataLayout::NCHW)))
+                                                                                                                     framework::dataset::make("DataLayout", DataLayout::NCHW)),
+                                                                                                                     ActivationFunctionsDataset))
 {
     validate(CLAccessor(_target), _reference, tolerance_f32);
 }
-TEST_SUITE_END()
+TEST_SUITE(Dilation)
+FIXTURE_DATA_TEST_CASE(RunSmall, CLDepthwiseConvolutionLayerFixture<float>, framework::DatasetMode::ALL,
+                       combine(combine(combine(combine(datasets::SmallDepthwiseDilatedConvolutionLayerDataset3x3(), depth_multipliers),
+                                               framework::dataset::make("DataType",
+                                                                        DataType::F32)),
+                                       framework::dataset::make("DataLayout", DataLayout::NCHW)),
+                               ActivationFunctionsDataset))
+{
+    validate(CLAccessor(_target), _reference, tolerance_f32);
+}
+FIXTURE_DATA_TEST_CASE(RunLarge, CLDepthwiseConvolutionLayerFixture<float>, framework::DatasetMode::NIGHTLY,
+                       combine(combine(combine(combine(datasets::LargeDepthwiseDilatedConvolutionLayerDataset3x3(),
+                                                       depth_multipliers),
+                                               framework::dataset::make("DataType",
+                                                                        DataType::F32)),
+                                       framework::dataset::make("DataLayout", DataLayout::NCHW)),
+                               ActivationFunctionsDataset))
+{
+    validate(CLAccessor(_target), _reference, tolerance_f32);
+}
+
+TEST_SUITE_END() // Dilation
+TEST_SUITE_END() // NCHW
 TEST_SUITE(NHWC)
 FIXTURE_DATA_TEST_CASE(RunSmall, CLDepthwiseConvolutionLayerFixture<float>, framework::DatasetMode::ALL,
-                       combine(combine(combine(datasets::SmallDepthwiseConvolutionLayerDataset3x3(),
-                                               depth_multipliers),
-                                       framework::dataset::make("DataType",
-                                                                DataType::F32)),
-                               framework::dataset::make("DataLayout", DataLayout::NHWC)))
+                       combine(combine(combine(combine(datasets::SmallDepthwiseConvolutionLayerDataset3x3(),
+                                                       depth_multipliers),
+                                               framework::dataset::make("DataType",
+                                                                        DataType::F32)),
+                                       framework::dataset::make("DataLayout", DataLayout::NHWC)),
+                               ActivationFunctionsDataset))
 {
     validate(CLAccessor(_target), _reference, tolerance_f32);
 }
-FIXTURE_DATA_TEST_CASE(RunLarge, CLDepthwiseConvolutionLayerFixture<float>, framework::DatasetMode::NIGHTLY, combine(combine(combine(datasets::LargeDepthwiseConvolutionLayerDataset3x3(),
+FIXTURE_DATA_TEST_CASE(RunLarge, CLDepthwiseConvolutionLayerFixture<float>, framework::DatasetMode::NIGHTLY, combine(combine(combine(combine(datasets::LargeDepthwiseConvolutionLayerDataset3x3(),
                                                                                                                      depth_multipliers),
                                                                                                                      framework::dataset::make("DataType",
                                                                                                                              DataType::F32)),
-                                                                                                                     framework::dataset::make("DataLayout", DataLayout::NHWC)))
+                                                                                                                     framework::dataset::make("DataLayout", DataLayout::NHWC)),
+                                                                                                                     ActivationFunctionsDataset))
 {
     validate(CLAccessor(_target), _reference, tolerance_f32);
 }
-TEST_SUITE_END()
-TEST_SUITE_END()
+TEST_SUITE(Dilation)
+
+FIXTURE_DATA_TEST_CASE(RunSmall, CLDepthwiseConvolutionLayerFixture<float>, framework::DatasetMode::ALL,
+                       combine(combine(combine(combine(datasets::SmallDepthwiseDilatedConvolutionLayerDataset3x3(),
+                                                       depth_multipliers),
+                                               framework::dataset::make("DataType",
+                                                                        DataType::F32)),
+                                       framework::dataset::make("DataLayout", DataLayout::NHWC)),
+                               ActivationFunctionsDataset))
+{
+    validate(CLAccessor(_target), _reference, tolerance_f32);
+}
+FIXTURE_DATA_TEST_CASE(RunLarge, CLDepthwiseConvolutionLayerFixture<float>, framework::DatasetMode::NIGHTLY,
+                       combine(combine(combine(combine(datasets::LargeDepthwiseDilatedConvolutionLayerDataset3x3(),
+                                                       depth_multipliers),
+                                               framework::dataset::make("DataType",
+                                                                        DataType::F32)),
+                                       framework::dataset::make("DataLayout", DataLayout::NHWC)),
+                               ActivationFunctionsDataset))
+{
+    validate(CLAccessor(_target), _reference, tolerance_f32);
+}
+TEST_SUITE_END() // Dilation
+TEST_SUITE_END() // NHWC
+TEST_SUITE_END() // W3x3
 
 TEST_SUITE(Generic)
-FIXTURE_DATA_TEST_CASE(RunSmall, CLDepthwiseConvolutionLayerFixture<float>, framework::DatasetMode::ALL, combine(combine(combine(datasets::SmallDepthwiseConvolutionLayerDataset(), depth_multipliers),
+FIXTURE_DATA_TEST_CASE(RunSmall, CLDepthwiseConvolutionLayerFixture<float>, framework::DatasetMode::ALL, combine(combine(combine(combine(datasets::SmallDepthwiseConvolutionLayerDataset(),
+                                                                                                                 depth_multipliers),
                                                                                                                  framework::dataset::make("DataType",
                                                                                                                          DataType::F32)),
-                                                                                                                 framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })))
+                                                                                                                 framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })),
+                                                                                                                 ActivationFunctionsDataset))
 {
     validate(CLAccessor(_target), _reference, tolerance_f32);
 }
-FIXTURE_DATA_TEST_CASE(RunLarge, CLDepthwiseConvolutionLayerFixture<float>, framework::DatasetMode::NIGHTLY, combine(combine(combine(datasets::LargeDepthwiseConvolutionLayerDataset(),
+FIXTURE_DATA_TEST_CASE(RunLarge, CLDepthwiseConvolutionLayerFixture<float>, framework::DatasetMode::NIGHTLY, combine(combine(combine(combine(datasets::LargeDepthwiseConvolutionLayerDataset(),
                                                                                                                      depth_multipliers),
                                                                                                                      framework::dataset::make("DataType",
                                                                                                                              DataType::F32)),
-                                                                                                                     framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })))
+                                                                                                                     framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })),
+                                                                                                                     ActivationFunctionsDataset))
 {
     validate(CLAccessor(_target), _reference, tolerance_f32);
 }
-TEST_SUITE_END()
-TEST_SUITE_END()
-TEST_SUITE_END()
+
+TEST_SUITE(Dilation)
+FIXTURE_DATA_TEST_CASE(RunSmall, CLDepthwiseConvolutionLayerFixture<float>, framework::DatasetMode::ALL, combine(combine(combine(combine(datasets::SmallDepthwiseDilatedConvolutionLayerDataset(),
+                                                                                                                 depth_multipliers),
+                                                                                                                 framework::dataset::make("DataType",
+                                                                                                                         DataType::F32)),
+                                                                                                                 framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })),
+                                                                                                                 ActivationFunctionsDataset))
+{
+    validate(CLAccessor(_target), _reference, tolerance_f32);
+}
+FIXTURE_DATA_TEST_CASE(RunLarge, CLDepthwiseConvolutionLayerFixture<float>, framework::DatasetMode::NIGHTLY,
+                       combine(combine(combine(combine(datasets::LargeDepthwiseDilatedConvolutionLayerDataset3x3(),
+                                                       depth_multipliers),
+                                               framework::dataset::make("DataType",
+                                                                        DataType::F32)),
+                                       framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })),
+                               ActivationFunctionsDataset))
+{
+    validate(CLAccessor(_target), _reference, tolerance_f32);
+}
+TEST_SUITE_END() // Dilation
+TEST_SUITE_END() // Generic
+TEST_SUITE_END() // FP32
+TEST_SUITE_END() // Float
 
 template <typename T>
 using CLDepthwiseConvolutionLayerQuantizedFixture = DepthwiseConvolutionLayerValidationQuantizedFixture<CLTensor, CLAccessor, CLDepthwiseConvolutionLayer, T>;
@@ -355,49 +557,97 @@
 TEST_SUITE(QASYMM8)
 TEST_SUITE(Generic)
 FIXTURE_DATA_TEST_CASE(RunSmall, CLDepthwiseConvolutionLayerQuantizedFixture<uint8_t>, framework::DatasetMode::PRECOMMIT,
-                       combine(combine(combine(combine(datasets::SmallDepthwiseConvolutionLayerDataset(),
-                                                       depth_multipliers),
-                                               framework::dataset::make("DataType", DataType::QASYMM8)),
-                                       framework::dataset::make("QuantizationInfo", { QuantizationInfo(0.5f, 10) })),
-                               framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })))
+                       combine(combine(combine(combine(combine(datasets::SmallDepthwiseConvolutionLayerDataset(),
+                                                               depth_multipliers),
+                                                       framework::dataset::make("DataType", DataType::QASYMM8)),
+                                               framework::dataset::make("QuantizationInfo", { QuantizationInfo(0.5f, 10) })),
+                                       framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })),
+                               ActivationFunctionsDataset))
 {
     validate(CLAccessor(_target), _reference, tolerance_qasymm8);
 }
 FIXTURE_DATA_TEST_CASE(RunLarge, CLDepthwiseConvolutionLayerQuantizedFixture<uint8_t>, framework::DatasetMode::NIGHTLY,
-                       combine(combine(combine(combine(datasets::LargeDepthwiseConvolutionLayerDataset(),
-                                                       depth_multipliers),
-                                               framework::dataset::make("DataType", DataType::QASYMM8)),
-                                       framework::dataset::make("QuantizationInfo", { QuantizationInfo(0.5f, 10) })),
-                               framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })))
+                       combine(combine(combine(combine(combine(datasets::LargeDepthwiseConvolutionLayerDataset(),
+                                                               depth_multipliers),
+                                                       framework::dataset::make("DataType", DataType::QASYMM8)),
+                                               framework::dataset::make("QuantizationInfo", { QuantizationInfo(0.5f, 10) })),
+                                       framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })),
+                               ActivationFunctionsDataset))
 {
     validate(CLAccessor(_target), _reference, tolerance_qasymm8);
 }
-TEST_SUITE_END()
+TEST_SUITE(Dilation)
+FIXTURE_DATA_TEST_CASE(RunSmall, CLDepthwiseConvolutionLayerQuantizedFixture<uint8_t>, framework::DatasetMode::PRECOMMIT,
+                       combine(combine(combine(combine(combine(datasets::SmallDepthwiseDilatedConvolutionLayerDataset(),
+                                                               depth_multipliers),
+                                                       framework::dataset::make("DataType", DataType::QASYMM8)),
+                                               framework::dataset::make("QuantizationInfo", { QuantizationInfo(0.5f, 10) })),
+                                       framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })),
+                               ActivationFunctionsDataset))
+{
+    validate(CLAccessor(_target), _reference, tolerance_qasymm8);
+}
+FIXTURE_DATA_TEST_CASE(RunLarge, CLDepthwiseConvolutionLayerQuantizedFixture<uint8_t>, framework::DatasetMode::NIGHTLY,
+                       combine(combine(combine(combine(combine(datasets::LargeDepthwiseDilatedConvolutionLayerDataset(),
+                                                               depth_multipliers),
+                                                       framework::dataset::make("DataType", DataType::QASYMM8)),
+                                               framework::dataset::make("QuantizationInfo", { QuantizationInfo(0.5f, 10) })),
+                                       framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })),
+                               ActivationFunctionsDataset))
+{
+    validate(CLAccessor(_target), _reference, tolerance_qasymm8);
+}
+TEST_SUITE_END() // Dilation
+TEST_SUITE_END() // Generic
 TEST_SUITE(W3x3)
 FIXTURE_DATA_TEST_CASE(RunSmall, CLDepthwiseConvolutionLayerQuantizedFixture<uint8_t>, framework::DatasetMode::PRECOMMIT,
-                       combine(combine(combine(combine(datasets::SmallDepthwiseConvolutionLayerDataset3x3(),
-                                                       depth_multipliers),
-                                               framework::dataset::make("DataType", DataType::QASYMM8)),
-                                       framework::dataset::make("QuantizationInfo", { QuantizationInfo(0.5f, 10) })),
-                               framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })))
+                       combine(combine(combine(combine(combine(datasets::SmallDepthwiseConvolutionLayerDataset3x3(),
+                                                               depth_multipliers),
+                                                       framework::dataset::make("DataType", DataType::QASYMM8)),
+                                               framework::dataset::make("QuantizationInfo", { QuantizationInfo(0.5f, 10) })),
+                                       framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })),
+                               ActivationFunctionsDataset))
 {
     validate(CLAccessor(_target), _reference, tolerance_qasymm8);
 }
 FIXTURE_DATA_TEST_CASE(RunLarge, CLDepthwiseConvolutionLayerQuantizedFixture<uint8_t>, framework::DatasetMode::NIGHTLY,
-                       combine(combine(combine(combine(datasets::LargeDepthwiseConvolutionLayerDataset3x3(),
-                                                       depth_multipliers),
-                                               framework::dataset::make("DataType", DataType::QASYMM8)),
-                                       framework::dataset::make("QuantizationInfo", { QuantizationInfo(0.5f, 10) })),
-                               framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })))
+                       combine(combine(combine(combine(combine(datasets::LargeDepthwiseConvolutionLayerDataset3x3(),
+                                                               depth_multipliers),
+                                                       framework::dataset::make("DataType", DataType::QASYMM8)),
+                                               framework::dataset::make("QuantizationInfo", { QuantizationInfo(0.5f, 10) })),
+                                       framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })),
+                               ActivationFunctionsDataset))
 {
     validate(CLAccessor(_target), _reference, tolerance_qasymm8);
 }
-TEST_SUITE_END()
-TEST_SUITE_END()
-TEST_SUITE_END()
+TEST_SUITE(Dilation)
+FIXTURE_DATA_TEST_CASE(RunSmall, CLDepthwiseConvolutionLayerQuantizedFixture<uint8_t>, framework::DatasetMode::PRECOMMIT,
+                       combine(combine(combine(combine(combine(datasets::SmallDepthwiseDilatedConvolutionLayerDataset3x3(),
+                                                               depth_multipliers),
+                                                       framework::dataset::make("DataType", DataType::QASYMM8)),
+                                               framework::dataset::make("QuantizationInfo", { QuantizationInfo(0.5f, 10) })),
+                                       framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })),
+                               ActivationFunctionsDataset))
+{
+    validate(CLAccessor(_target), _reference, tolerance_qasymm8);
+}
+FIXTURE_DATA_TEST_CASE(RunLarge, CLDepthwiseConvolutionLayerQuantizedFixture<uint8_t>, framework::DatasetMode::NIGHTLY,
+                       combine(combine(combine(combine(combine(datasets::LargeDepthwiseDilatedConvolutionLayerDataset3x3(),
+                                                               depth_multipliers),
+                                                       framework::dataset::make("DataType", DataType::QASYMM8)),
+                                               framework::dataset::make("QuantizationInfo", { QuantizationInfo(0.5f, 10) })),
+                                       framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })),
+                               ActivationFunctionsDataset))
+{
+    validate(CLAccessor(_target), _reference, tolerance_qasymm8);
+}
+TEST_SUITE_END() // Dilation
+TEST_SUITE_END() // W3x3
+TEST_SUITE_END() // QASYMM8
+TEST_SUITE_END() // Quantized
 
-TEST_SUITE_END()
-TEST_SUITE_END()
+TEST_SUITE_END() // DepthwiseConvolutionLayer
+TEST_SUITE_END() // CL
 } // namespace validation
 } // namespace test
 } // namespace arm_compute
diff --git a/tests/validation/CL/DequantizationLayer.cpp b/tests/validation/CL/DequantizationLayer.cpp
index 5303566..b1b0d81 100644
--- a/tests/validation/CL/DequantizationLayer.cpp
+++ b/tests/validation/CL/DequantizationLayer.cpp
@@ -40,107 +40,94 @@
 {
 namespace validation
 {
-namespace
-{
-const auto DequantizationShapes = concat(datasets::Small3DShapes(),
-                                         datasets::Small4DShapes());
-} // namespace
-
 TEST_SUITE(CL)
 TEST_SUITE(DequantizationLayer)
 
 // *INDENT-OFF*
 // clang-format off
-DATA_TEST_CASE(Validate, framework::DatasetMode::ALL, zip(zip(zip(
-               framework::dataset::make("InputInfo", { TensorInfo(TensorShape(16U, 16U, 16U, 5U), 1, DataType::F32), // Wrong input data type
-                                                       TensorInfo(TensorShape(16U, 5U, 16U), 1, DataType::U8),       // Invalid shape
-                                                       TensorInfo(TensorShape(16U, 16U, 16U, 5U), 1, DataType::U8),  // Wrong output data type
-                                                       TensorInfo(TensorShape(16U, 16U, 2U, 5U), 1, DataType::U8),   // Missmatching shapes
-                                                       TensorInfo(TensorShape(17U, 16U, 16U, 5U), 1, DataType::U8),  // Shrink window
-                                                       TensorInfo(TensorShape(16U, 16U, 16U, 5U), 1, DataType::U8),  // Valid
+DATA_TEST_CASE(Validate, framework::DatasetMode::ALL, zip(zip(
+               framework::dataset::make("InputInfo", { TensorInfo(TensorShape(16U, 16U, 16U, 5U), 1, DataType::F32),      // Wrong input data type
+                                                       TensorInfo(TensorShape(16U, 16U, 16U, 5U), 1, DataType::QASYMM8),  // Wrong output data type
+                                                       TensorInfo(TensorShape(16U, 16U, 2U, 5U), 1, DataType::QASYMM8),   // Missmatching shapes
+                                                       TensorInfo(TensorShape(17U, 16U, 16U, 5U), 1, DataType::QASYMM8),  // Valid
+                                                       TensorInfo(TensorShape(16U, 16U, 16U, 5U), 1, DataType::QASYMM8),  // Valid
                                                      }),
                framework::dataset::make("OutputInfo",{ TensorInfo(TensorShape(16U, 16U, 16U, 5U), 1, DataType::F32),
-                                                       TensorInfo(TensorShape(16U, 5U, 16U), 1, DataType::U8),
                                                        TensorInfo(TensorShape(16U, 16U, 16U, 5U), 1, DataType::U8),
                                                        TensorInfo(TensorShape(16U, 16U, 16U, 5U), 1, DataType::F32),
                                                        TensorInfo(TensorShape(17U, 16U, 16U, 5U), 1, DataType::F32),
                                                        TensorInfo(TensorShape(16U, 16U, 16U, 5U), 1, DataType::F32),
                                                      })),
-               framework::dataset::make("MinMax",{ TensorInfo(TensorShape(2U), 1, DataType::F32),
-                                                       TensorInfo(TensorShape(2U), 1, DataType::U8),
-                                                       TensorInfo(TensorShape(2U), 1, DataType::F32),
-                                                       TensorInfo(TensorShape(2U), 1, DataType::F32),
-                                                       TensorInfo(TensorShape(2U), 1, DataType::U8),
-                                                       TensorInfo(TensorShape(2U), 1, DataType::U8),
-                                                     })),
-               framework::dataset::make("Expected", { false, false, false, false, false, true})),
-               input_info, output_info, min_max, expected)
+               framework::dataset::make("Expected", { false, false, false, true, true})),
+               input_info, output_info, expected)
 {
-    ARM_COMPUTE_EXPECT(bool(CLDequantizationLayer::validate(&input_info.clone()->set_is_resizable(false), &output_info.clone()->set_is_resizable(false), &min_max.clone()->set_is_resizable(false))) == expected, framework::LogLevel::ERRORS);
+    ARM_COMPUTE_EXPECT(bool(CLDequantizationLayer::validate(&input_info.clone()->set_is_resizable(false), &output_info.clone()->set_is_resizable(false))) == expected, framework::LogLevel::ERRORS);
 }
 // clang-format on
 // *INDENT-ON*
 
-DATA_TEST_CASE(Configuration, framework::DatasetMode::ALL, combine(DequantizationShapes, framework::dataset::make("DataType", DataType::U8)), shape, data_type)
+DATA_TEST_CASE(Configuration,
+               framework::DatasetMode::ALL,
+               combine(datasets::SmallShapes(), framework::dataset::make("DataType", { DataType::F16, DataType::F32 })),
+               shape, data_type)
 {
-    TensorShape shape_min_max = shape;
-    shape_min_max.set(Window::DimX, 2);
-
-    // Remove Y and Z dimensions and keep the batches
-    shape_min_max.remove_dimension(1);
-    shape_min_max.remove_dimension(1);
-
     // Create tensors
-    CLTensor src     = create_tensor<CLTensor>(shape, data_type);
-    CLTensor dst     = create_tensor<CLTensor>(shape, DataType::F32);
-    CLTensor min_max = create_tensor<CLTensor>(shape_min_max, DataType::F32);
+    CLTensor src = create_tensor<CLTensor>(shape, DataType::QASYMM8, 1, QuantizationInfo(0.5f, -10));
+    CLTensor dst = create_tensor<CLTensor>(shape, data_type);
 
     ARM_COMPUTE_EXPECT(src.info()->is_resizable(), framework::LogLevel::ERRORS);
     ARM_COMPUTE_EXPECT(dst.info()->is_resizable(), framework::LogLevel::ERRORS);
-    ARM_COMPUTE_EXPECT(min_max.info()->is_resizable(), framework::LogLevel::ERRORS);
 
     // Create and configure function
     CLDequantizationLayer dequant_layer;
-    dequant_layer.configure(&src, &dst, &min_max);
+    dequant_layer.configure(&src, &dst);
 
     // Validate valid region
     const ValidRegion valid_region = shape_to_valid_region(shape);
     validate(src.info()->valid_region(), valid_region);
     validate(dst.info()->valid_region(), valid_region);
 
-    // Validate valid region of min_max tensor
-    const ValidRegion valid_region_min_max = shape_to_valid_region(shape_min_max);
-    validate(min_max.info()->valid_region(), valid_region_min_max);
-
     // Validate padding
-    const PaddingSize padding = PaddingCalculator(shape.x(), 4).required_padding();
-    validate(src.info()->padding(), padding);
-    validate(dst.info()->padding(), padding);
-
-    // Validate padding of min_max tensor
-    const PaddingSize padding_min_max = PaddingCalculator(shape_min_max.x(), 2).required_padding();
-    validate(min_max.info()->padding(), padding_min_max);
+    validate(src.info()->padding(), PaddingSize());
+    validate(dst.info()->padding(), PaddingSize());
 }
 
 template <typename T>
 using CLDequantizationLayerFixture = DequantizationValidationFixture<CLTensor, CLAccessor, CLDequantizationLayer, T>;
 
-TEST_SUITE(Integer)
-TEST_SUITE(U8)
-FIXTURE_DATA_TEST_CASE(RunSmall, CLDequantizationLayerFixture<uint8_t>, framework::DatasetMode::PRECOMMIT, combine(concat(datasets::Small3DShapes(), datasets::Small4DShapes()),
-                                                                                                                   framework::dataset::make("DataType", DataType::U8)))
+TEST_SUITE(FP16)
+FIXTURE_DATA_TEST_CASE(RunSmall, CLDequantizationLayerFixture<half>, framework::DatasetMode::PRECOMMIT, combine(combine(datasets::SmallShapes(),
+                                                                                                                        framework::dataset::make("DataType", DataType::F16)),
+                                                                                                                framework::dataset::make("QuantizationInfo", { QuantizationInfo(0.1f, 128.0f) })))
 {
     // Validate output
     validate(CLAccessor(_target), _reference);
 }
-FIXTURE_DATA_TEST_CASE(RunLarge, CLDequantizationLayerFixture<uint8_t>, framework::DatasetMode::NIGHTLY, combine(concat(datasets::Large3DShapes(), datasets::Large4DShapes()),
-                                                                                                                 framework::dataset::make("DataType", DataType::U8)))
+FIXTURE_DATA_TEST_CASE(RunLarge, CLDequantizationLayerFixture<half>, framework::DatasetMode::NIGHTLY, combine(combine(datasets::LargeShapes(),
+                                                                                                                      framework::dataset::make("DataType", DataType::F16)),
+                                                                                                              framework::dataset::make("QuantizationInfo", { QuantizationInfo(0.1f, 128.0f) })))
 {
     // Validate output
     validate(CLAccessor(_target), _reference);
 }
-TEST_SUITE_END() // U8
-TEST_SUITE_END() // Integer
+TEST_SUITE_END() // FP16
+
+TEST_SUITE(FP32)
+FIXTURE_DATA_TEST_CASE(RunSmall, CLDequantizationLayerFixture<float>, framework::DatasetMode::PRECOMMIT, combine(combine(datasets::SmallShapes(),
+                                                                                                                 framework::dataset::make("DataType", DataType::F32)),
+                                                                                                                 framework::dataset::make("QuantizationInfo", { QuantizationInfo(0.1f, 128.0f) })))
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference);
+}
+FIXTURE_DATA_TEST_CASE(RunLarge, CLDequantizationLayerFixture<float>, framework::DatasetMode::NIGHTLY, combine(combine(datasets::LargeShapes(),
+                                                                                                                       framework::dataset::make("DataType", DataType::F32)),
+                                                                                                               framework::dataset::make("QuantizationInfo", { QuantizationInfo(0.1f, 128.0f) })))
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference);
+}
+TEST_SUITE_END() // FP32
 
 TEST_SUITE_END() // DequantizationLayer
 TEST_SUITE_END() // CL
diff --git a/tests/validation/CL/FFT.cpp b/tests/validation/CL/FFT.cpp
new file mode 100644
index 0000000..9fdd85b
--- /dev/null
+++ b/tests/validation/CL/FFT.cpp
@@ -0,0 +1,228 @@
+/*
+ * Copyright (c) 2019 ARM Limited.
+ *
+ * SPDX-License-Identifier: MIT
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#include "arm_compute/core/Types.h"
+#include "arm_compute/runtime/CL/CLTensor.h"
+#include "arm_compute/runtime/CL/functions/CLFFT1D.h"
+#include "arm_compute/runtime/CL/functions/CLFFT2D.h"
+#include "arm_compute/runtime/CL/functions/CLFFTConvolutionLayer.h"
+#include "tests/CL/CLAccessor.h"
+#include "tests/datasets/SmallConvolutionLayerDataset.h"
+#include "tests/framework/Asserts.h"
+#include "tests/framework/Macros.h"
+#include "tests/framework/datasets/Datasets.h"
+#include "tests/validation/Validation.h"
+#include "tests/validation/fixtures/FFTFixture.h"
+
+namespace arm_compute
+{
+namespace test
+{
+namespace validation
+{
+namespace
+{
+const auto data_types = framework::dataset::make("DataType", { DataType::F32 });
+const auto shapes_1d  = framework::dataset::make("TensorShape", { TensorShape(2U, 2U, 3U), TensorShape(3U, 2U, 3U),
+                                                                  TensorShape(4U, 2U, 3U), TensorShape(5U, 2U, 3U),
+                                                                  TensorShape(7U, 2U, 3U), TensorShape(8U, 2U, 3U),
+                                                                  TensorShape(9U, 2U, 3U), TensorShape(25U, 2U, 3U),
+                                                                  TensorShape(49U, 2U, 3U), TensorShape(64U, 2U, 3U),
+                                                                  TensorShape(16U, 2U, 3U), TensorShape(32U, 2U, 3U),
+                                                                  TensorShape(96U, 2U, 2U)
+                                                                });
+const auto shapes_2d = framework::dataset::make("TensorShape", { TensorShape(2U, 2U, 3U), TensorShape(3U, 6U, 3U),
+                                                                 TensorShape(4U, 5U, 3U), TensorShape(5U, 7U, 3U),
+                                                                 TensorShape(7U, 25U, 3U), TensorShape(8U, 2U, 3U),
+                                                                 TensorShape(9U, 16U, 3U), TensorShape(25U, 32U, 3U),
+                                                                 TensorShape(192U, 128U, 2U)
+                                                               });
+
+const auto ActivationFunctionsSmallDataset = framework::dataset::make("ActivationInfo",
+{
+    ActivationLayerInfo(),
+    ActivationLayerInfo(ActivationLayerInfo::ActivationFunction::LU_BOUNDED_RELU, 0.5f)
+});
+
+RelativeTolerance<float> tolerance_f32(0.1f);   /**< Relative tolerance value for FP32 */
+constexpr float          tolerance_num = 0.07f; /**< Tolerance number */
+
+} // namespace
+TEST_SUITE(CL)
+TEST_SUITE(FFT1D)
+
+DATA_TEST_CASE(Configuration, framework::DatasetMode::ALL, combine(shapes_1d, data_types),
+               shape, data_type)
+{
+    // Create tensors
+    CLTensor src = create_tensor<CLTensor>(shape, data_type, 2);
+    CLTensor dst = create_tensor<CLTensor>(shape, data_type, 2);
+
+    ARM_COMPUTE_EXPECT(src.info()->is_resizable(), framework::LogLevel::ERRORS);
+    ARM_COMPUTE_EXPECT(dst.info()->is_resizable(), framework::LogLevel::ERRORS);
+
+    // Create and configure function
+    CLFFT1D fft1d;
+    fft1d.configure(&src, &dst, FFT1DInfo());
+
+    // Validate valid region
+    const ValidRegion valid_region = shape_to_valid_region(shape);
+    validate(src.info()->valid_region(), valid_region);
+    validate(dst.info()->valid_region(), valid_region);
+
+    // Validate padding
+    validate(src.info()->padding(), PaddingSize());
+    validate(dst.info()->padding(), PaddingSize());
+}
+
+// *INDENT-OFF*
+// clang-format off
+DATA_TEST_CASE(Validate, framework::DatasetMode::ALL, zip(zip(zip(
+        framework::dataset::make("InputInfo", { TensorInfo(TensorShape(32U, 13U, 2U), 2, DataType::F32), // Mismatching data types
+                                                TensorInfo(TensorShape(32U, 13U, 2U), 2, DataType::F32), // Mismatching shapes
+                                                TensorInfo(TensorShape(32U, 13U, 2U), 3, DataType::F32), // Invalid channels
+                                                TensorInfo(TensorShape(32U, 13U, 2U), 2, DataType::F32), // Unsupported axis
+                                                TensorInfo(TensorShape(11U, 13U, 2U), 2, DataType::F32), // Undecomposable FFT
+                                                TensorInfo(TensorShape(25U, 13U, 2U), 2, DataType::F32),
+        }),
+        framework::dataset::make("OutputInfo",{ TensorInfo(TensorShape(32U, 13U, 2U), 2, DataType::F16),
+                                                TensorInfo(TensorShape(16U, 13U, 2U), 2, DataType::F32),
+                                                TensorInfo(TensorShape(32U, 13U, 2U), 2, DataType::F32),
+                                                TensorInfo(TensorShape(32U, 13U, 2U), 2, DataType::F32),
+                                                TensorInfo(TensorShape(11U, 13U, 2U), 2, DataType::F32),
+                                                TensorInfo(TensorShape(25U, 13U, 2U), 2, DataType::F32),
+        })),
+        framework::dataset::make("Axis", { 0, 0, 0, 2, 0, 0 })),
+        framework::dataset::make("Expected", { false, false, false, false, false, true })),
+        input_info, output_info, axis, expected)
+{
+    FFT1DInfo desc;
+    desc.axis = axis;
+    const Status s = CLFFT1D::validate(&input_info.clone()->set_is_resizable(false), &output_info.clone()->set_is_resizable(false), desc);
+    ARM_COMPUTE_EXPECT(bool(s) == expected, framework::LogLevel::ERRORS);
+}
+// clang-format on
+// *INDENT-ON*
+
+template <typename T>
+using CLFFT1DFixture = FFTValidationFixture<CLTensor, CLAccessor, CLFFT1D, FFT1DInfo, T>;
+
+TEST_SUITE(Float)
+TEST_SUITE(FP32)
+FIXTURE_DATA_TEST_CASE(RunSmall, CLFFT1DFixture<float>, framework::DatasetMode::ALL, combine(shapes_1d, framework::dataset::make("DataType", DataType::F32)))
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference, tolerance_f32, tolerance_num);
+}
+TEST_SUITE_END() // FP32
+TEST_SUITE_END() // Float
+
+TEST_SUITE_END() // FFT1D
+
+TEST_SUITE(FFT2D)
+
+DATA_TEST_CASE(Configuration, framework::DatasetMode::ALL, combine(shapes_2d, data_types),
+               shape, data_type)
+{
+    // Create tensors
+    CLTensor src = create_tensor<CLTensor>(shape, data_type, 2);
+    CLTensor dst = create_tensor<CLTensor>(shape, data_type, 2);
+
+    ARM_COMPUTE_EXPECT(src.info()->is_resizable(), framework::LogLevel::ERRORS);
+    ARM_COMPUTE_EXPECT(dst.info()->is_resizable(), framework::LogLevel::ERRORS);
+
+    // Create and configure function
+    CLFFT2D fft2d;
+    fft2d.configure(&src, &dst, FFT2DInfo());
+
+    // Validate valid region
+    const ValidRegion valid_region = shape_to_valid_region(shape);
+    validate(src.info()->valid_region(), valid_region);
+    validate(dst.info()->valid_region(), valid_region);
+
+    // Validate padding
+    validate(src.info()->padding(), PaddingSize());
+    validate(dst.info()->padding(), PaddingSize());
+}
+
+// *INDENT-OFF*
+// clang-format off
+DATA_TEST_CASE(Validate, framework::DatasetMode::ALL, zip(zip(
+        framework::dataset::make("InputInfo", { TensorInfo(TensorShape(32U, 25U, 2U), 2, DataType::F32), // Mismatching data types
+                                                TensorInfo(TensorShape(32U, 25U, 2U), 2, DataType::F32), // Mismatching shapes
+                                                TensorInfo(TensorShape(32U, 25U, 2U), 3, DataType::F32), // Invalid channels
+                                                TensorInfo(TensorShape(32U, 13U, 2U), 2, DataType::F32), // Undecomposable FFT
+                                                TensorInfo(TensorShape(32U, 25U, 2U), 2, DataType::F32),
+        }),
+        framework::dataset::make("OutputInfo",{ TensorInfo(TensorShape(32U, 25U, 2U), 2, DataType::F16),
+                                                TensorInfo(TensorShape(16U, 25U, 2U), 2, DataType::F32),
+                                                TensorInfo(TensorShape(32U, 25U, 2U), 1, DataType::F32),
+                                                TensorInfo(TensorShape(32U, 13U, 2U), 2, DataType::F32),
+                                                TensorInfo(TensorShape(32U, 25U, 2U), 2, DataType::F32),
+        })),
+        framework::dataset::make("Expected", { false, false, false, false, true })),
+               input_info, output_info, expected)
+{
+    const Status s = CLFFT2D::validate(&input_info.clone()->set_is_resizable(false), &output_info.clone()->set_is_resizable(false), FFT2DInfo());
+    ARM_COMPUTE_EXPECT(bool(s) == expected, framework::LogLevel::ERRORS);
+}
+// clang-format on
+// *INDENT-ON*
+
+template <typename T>
+using CLFFT2DFixture = FFTValidationFixture<CLTensor, CLAccessor, CLFFT2D, FFT2DInfo, T>;
+
+TEST_SUITE(Float)
+TEST_SUITE(FP32)
+FIXTURE_DATA_TEST_CASE(RunSmall, CLFFT2DFixture<float>, framework::DatasetMode::ALL, combine(shapes_2d, framework::dataset::make("DataType", DataType::F32)))
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference, tolerance_f32, tolerance_num);
+}
+TEST_SUITE_END() // FP32
+TEST_SUITE_END() // Float
+TEST_SUITE_END() // FFT2D
+
+TEST_SUITE(FFTConvolutionLayer)
+
+template <typename T>
+using CLFFTConvolutionLayerFixture = FFTConvolutionValidationFixture<CLTensor, CLAccessor, CLFFTConvolutionLayer, T>;
+
+TEST_SUITE(Float)
+TEST_SUITE(FP32)
+FIXTURE_DATA_TEST_CASE(RunSmall, CLFFTConvolutionLayerFixture<float>, framework::DatasetMode::PRECOMMIT, combine(combine(combine(datasets::SmallFFTConvolutionLayerDataset(),
+                                                                                                                 framework::dataset::make("DataType", DataType::F32)),
+                                                                                                                 framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })),
+                                                                                                                 ActivationFunctionsSmallDataset))
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference, tolerance_f32, tolerance_num);
+}
+TEST_SUITE_END() // FP32
+TEST_SUITE_END() // Float
+TEST_SUITE_END() // FFTConvolutionLayer
+
+TEST_SUITE_END() // CL
+} // namespace validation
+} // namespace test
+} // namespace arm_compute
diff --git a/tests/validation/CL/GEMMLowp.cpp b/tests/validation/CL/GEMMLowp.cpp
index 08641db..efefbd6 100644
--- a/tests/validation/CL/GEMMLowp.cpp
+++ b/tests/validation/CL/GEMMLowp.cpp
@@ -28,6 +28,7 @@
 #include "arm_compute/runtime/CL/functions/CLGEMMLowpOutputStage.h"
 #include "tests/CL/CLAccessor.h"
 #include "tests/PaddingCalculator.h"
+#include "tests/datasets/GEMMLowpFusedOffsetOutputDataset.h"
 #include "tests/datasets/LargeGEMMLowpDataset.h"
 #include "tests/datasets/ShapeDatasets.h"
 #include "tests/datasets/SmallGEMMLowpDataset.h"
@@ -83,6 +84,21 @@
     validate(CLAccessor(_target), _reference);
 }
 
+using CLGEMMLowpMatrixMultiplyCoreFusedOffsetOutputFixture = GEMMLowpMatrixMultiplyCoreFusedOffsetOutputValidationFixture<CLTensor, CLAccessor, CLGEMMLowpMatrixMultiplyCore>;
+TEST_SUITE(FusedOffsetOutput)
+FIXTURE_DATA_TEST_CASE(RunSmall, CLGEMMLowpMatrixMultiplyCoreFusedOffsetOutputFixture, framework::DatasetMode::ALL, datasets::SmallGEMMLowpFusedOffsetOutputDataset())
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference);
+}
+
+FIXTURE_DATA_TEST_CASE(RunLarge, CLGEMMLowpMatrixMultiplyCoreFusedOffsetOutputFixture, framework::DatasetMode::NIGHTLY, datasets::LargeGEMMLowpFusedOffsetOutputDataset())
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference);
+}
+TEST_SUITE_END() // FusedOffsetOutput
+
 TEST_SUITE(Output3D)
 using CLGEMMLowpMatrixMultiplyCoreOutput3DFixture = GEMMLowpMatrixMultiplyCoreValidationFixture<CLTensor, CLAccessor, CLGEMMLowpMatrixMultiplyCore, false, true>;
 FIXTURE_DATA_TEST_CASE(RunSmall, CLGEMMLowpMatrixMultiplyCoreOutput3DFixture, framework::DatasetMode::PRECOMMIT, datasets::SmallGEMMLowpOutput3DDataset())
diff --git a/tests/validation/CL/GEMMLowpMatrixMultiplyReshaped.cpp b/tests/validation/CL/GEMMLowpMatrixMultiplyReshaped.cpp
index 62b0d02..60b92bd 100644
--- a/tests/validation/CL/GEMMLowpMatrixMultiplyReshaped.cpp
+++ b/tests/validation/CL/GEMMLowpMatrixMultiplyReshaped.cpp
@@ -192,7 +192,7 @@
     // Validate output
     validate(CLAccessor(_target), _reference);
 }
-TEST_SUITE_END() // GEMMLowpMatrixMulipltyReshaped
+TEST_SUITE_END() // GEMMLowpMatrixMultiplyReshaped
 TEST_SUITE_END() // CL
 } // namespace validation
 } // namespace test
diff --git a/tests/validation/CL/GEMMLowpMatrixMultiplyReshapedOnlyRHS.cpp b/tests/validation/CL/GEMMLowpMatrixMultiplyReshapedOnlyRHS.cpp
new file mode 100644
index 0000000..a907c5b
--- /dev/null
+++ b/tests/validation/CL/GEMMLowpMatrixMultiplyReshapedOnlyRHS.cpp
@@ -0,0 +1,248 @@
+/*
+ * Copyright (c) 2019 ARM Limited.
+ *
+ * SPDX-License-Identifier: MIT
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#include "arm_compute/core/CL/kernels/CLGEMMLowpMatrixMultiplyReshapedOnlyRHSKernel.h"
+#include "arm_compute/core/CL/kernels/CLGEMMReshapeRHSMatrixKernel.h"
+#include "arm_compute/core/Types.h"
+#include "arm_compute/core/utils/misc/ShapeCalculator.h"
+#include "arm_compute/runtime/CL/CLTensor.h"
+#include "arm_compute/runtime/CL/CLTensorAllocator.h"
+#include "tests/CL/CLAccessor.h"
+#include "tests/CL/Helper.h"
+#include "tests/PaddingCalculator.h"
+#include "tests/datasets/ShapeDatasets.h"
+#include "tests/framework/Asserts.h"
+#include "tests/framework/Macros.h"
+#include "tests/framework/datasets/Datasets.h"
+#include "tests/validation/Validation.h"
+#include "tests/validation/fixtures/GEMMLowpFixture.h"
+
+namespace arm_compute
+{
+namespace test
+{
+namespace validation
+{
+using namespace arm_compute::misc::shape_calculator;
+
+// Create function for CLGEMMReshapeRHSMatrixKernel
+using CLGEMMReshapeRHSMatrix = CLSynthetizeFunction<CLGEMMReshapeRHSMatrixKernel>;
+
+// Create function for CLGEMMLowpMatrixMultiplyReshapedOnlyRHSKernel
+using CLGEMMLowpMatrixMultiplyReshapedOnlyRHS = CLSynthetizeFunction<CLGEMMLowpMatrixMultiplyReshapedOnlyRHSKernel>;
+
+// Fixture for CLGEMMLowpMatrixMultiplyReshapedOnlyRHS
+using CLGEMMLowpMatrixMultiplyReshapedOnlyRHSFixture = GEMMLowpMatrixMultiplyReshapedOnlyRHSValidationFixture<CLTensor, CLAccessor, CLGEMMReshapeRHSMatrix, CLGEMMLowpMatrixMultiplyReshapedOnlyRHS>;
+
+// Fixture for CLGEMMLowpMatrixMultiplyReshapedOnlyRHS3D
+using CLGEMMLowpMatrixMultiplyReshapedOnlyRHS3DFixture =
+    GEMMLowpMatrixMultiplyReshapedOnlyRHS3DValidationFixture<CLTensor, CLAccessor, CLGEMMReshapeRHSMatrix, CLGEMMLowpMatrixMultiplyReshapedOnlyRHS>;
+
+namespace
+{
+// *INDENT-OFF*
+// clang-format off
+
+/** M values to test */
+const auto m_values = framework::dataset::make("M", 37);
+
+/** M_W values to test */
+const auto m_w_values = framework::dataset::make("M_W", 5);
+
+/** M_H values to test */
+const auto m_h_values = framework::dataset::make("M_H", 7);
+
+/** N values to test */
+const auto n_values = framework::dataset::make("N", 51);
+
+/** K values to test */
+const auto k_values = framework::dataset::make("K", 23);
+
+/** Batch size values to test */
+const auto b_values = framework::dataset::make("batch_size", 1, 3);
+
+/** M0 values to test - Precommit */
+const auto m0_values_precommit = framework::dataset::make("M0", {4, 6});
+
+/** N0 values to test - Precommit */
+const auto n0_values_precommit = framework::dataset::make("N0", { 2, 4 });
+
+/** K0 values to test - Precommit */
+const auto k0_values_precommit = framework::dataset::make("K0", { 4 });
+
+/** H0 values to test - Precommit */
+const auto h0_values_precommit = framework::dataset::make("H0", 1, 3);
+
+/** M0 values to test - Nightly */
+const auto m0_values_nightly = framework::dataset::make("M0", 2, 8);
+
+/** N0 values to test - Nightly */
+const auto n0_values_nightly = framework::dataset::make("N0", { 2, 3, 4, 8 });
+
+/** K0 values to test - Nightly */
+const auto k0_values_nightly = framework::dataset::make("K0", { 2, 3, 4, 8 });
+
+/** H0 values to test - Nightly */
+const auto h0_values_nightly = framework::dataset::make("H0", 1, 4);
+
+/** Interleave values to test with RHS matrix */
+const auto i_values_rhs = framework::dataset::make("interleave_rhs", { true, false });
+
+/** Transpose values to test with RHS matrix */
+const auto t_values_rhs = framework::dataset::make("transpose_rhs", { true });
+
+/** Configuration test */
+void validate_configuration(unsigned int m_value, unsigned int n_value, unsigned int k_value, unsigned int b_value, unsigned int m0_value, unsigned int n0_value, unsigned int k0_value, unsigned int h0_value, bool i_value_rhs)
+{
+    const unsigned int M = m_value;
+    const unsigned int N = n_value;
+    const unsigned int K = k_value;
+
+    GEMMLHSMatrixInfo lhs_info;
+    lhs_info.m0         = m0_value;
+    lhs_info.k0         = k0_value;
+
+    GEMMRHSMatrixInfo rhs_info;
+    rhs_info.n0         = n0_value;
+    rhs_info.k0         = k0_value;
+    rhs_info.h0         = h0_value;
+    rhs_info.interleave = i_value_rhs;
+    rhs_info.transpose  = true;
+
+    GEMMReshapeInfo gemm_info(M, N, K);
+
+    const TensorShape lhs_shape(K, M, b_value);
+    const TensorShape rhs_shape(N, K, b_value);
+    const TensorShape rhs_shape_reshaped = compute_rhs_reshaped_shape(TensorInfo(rhs_shape, 1, DataType::QASYMM8),
+                                                                      rhs_info);
+
+    const TensorShape dst_shape = compute_mm_shape(TensorInfo(lhs_shape, 1, DataType::QASYMM8),
+                                                   TensorInfo(rhs_shape_reshaped, 1, DataType::QASYMM8),
+                                                   gemm_info);
+
+    // Create tensors
+    CLTensor lhs          = create_tensor<CLTensor>(lhs_shape, DataType::QASYMM8);
+    CLTensor rhs_reshaped = create_tensor<CLTensor>(rhs_shape_reshaped, DataType::QASYMM8);
+    CLTensor dst          = create_tensor<CLTensor>(dst_shape, DataType::S32);
+
+    ARM_COMPUTE_EXPECT(lhs.info()->is_resizable(), framework::LogLevel::ERRORS);
+    ARM_COMPUTE_EXPECT(rhs_reshaped.info()->is_resizable(), framework::LogLevel::ERRORS);
+    ARM_COMPUTE_EXPECT(dst.info()->is_resizable(), framework::LogLevel::ERRORS);
+
+    // Create and configure function
+    CLGEMMLowpMatrixMultiplyReshapedOnlyRHS gemm;
+    gemm.configure(&lhs, &rhs_reshaped, &dst, lhs_info, rhs_info, gemm_info);
+}
+} // namespace
+
+TEST_SUITE(CL)
+TEST_SUITE(GEMMLowpMatrixMultiplyReshapedOnlyRHS)
+DATA_TEST_CASE(Configuration, framework::DatasetMode::ALL, combine(combine(combine(combine(combine(combine(combine(combine(
+                                                                   m_values,
+                                                                   n_values),
+                                                                   k_values),
+                                                                   framework::dataset::make("batch_size", 1)),
+                                                                   m0_values_precommit),
+                                                                   n0_values_precommit),
+                                                                   k0_values_precommit),
+                                                                   h0_values_precommit),
+                                                                   i_values_rhs),
+m_value, n_value, k_value, b_value, m0_value, n0_value, k0_value, h0_value, i_value_rhs)
+{
+    validate_configuration(m_value, n_value, k_value, b_value, m0_value, n0_value, k0_value, h0_value, i_value_rhs);
+}
+
+FIXTURE_DATA_TEST_CASE(RunSmall, CLGEMMLowpMatrixMultiplyReshapedOnlyRHSFixture, framework::DatasetMode::ALL,
+                combine(combine(combine(combine(combine(combine(combine(combine(combine(
+                                                                   m_values,
+                                                                   n_values),
+                                                                   k_values),
+                                                                   b_values),
+                                                                   m0_values_precommit),
+                                                                   n0_values_precommit),
+                                                                   k0_values_precommit),
+                                                                   h0_values_precommit),
+                                                                   i_values_rhs),
+                                                                   t_values_rhs))
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference);
+}
+
+FIXTURE_DATA_TEST_CASE(RunLarge, CLGEMMLowpMatrixMultiplyReshapedOnlyRHSFixture, framework::DatasetMode::NIGHTLY,
+                combine(combine(combine(combine(combine(combine(combine(combine(combine(
+                                                                   m_values,
+                                                                   n_values),
+                                                                   k_values),
+                                                                   b_values),
+                                                                   m0_values_nightly),
+                                                                   n0_values_nightly),
+                                                                   k0_values_nightly),
+                                                                   h0_values_nightly),
+                                                                   i_values_rhs),
+                                                                   t_values_rhs))
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference);
+}
+
+FIXTURE_DATA_TEST_CASE(RunSmall3D, CLGEMMLowpMatrixMultiplyReshapedOnlyRHS3DFixture, framework::DatasetMode::ALL,
+                combine(combine(combine(combine(combine(combine(combine(combine(combine(combine(
+                                                                   m_w_values,
+                                                                   m_h_values),
+                                                                   n_values),
+                                                                   k_values),
+                                                                   b_values),
+                                                                   m0_values_precommit),
+                                                                   n0_values_precommit),
+                                                                   k0_values_precommit),
+                                                                   h0_values_precommit),
+                                                                   i_values_rhs),
+                                                                   t_values_rhs))
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference);
+}
+
+FIXTURE_DATA_TEST_CASE(RunLarge3D, CLGEMMLowpMatrixMultiplyReshapedOnlyRHS3DFixture, framework::DatasetMode::NIGHTLY,
+                combine(combine(combine(combine(combine(combine(combine(combine(combine(combine(
+                                                                   m_w_values,
+                                                                   m_h_values),
+                                                                   n_values),
+                                                                   k_values),
+                                                                   b_values),
+                                                                   m0_values_nightly),
+                                                                   n0_values_nightly),
+                                                                   k0_values_nightly),
+                                                                   h0_values_nightly),
+                                                                   i_values_rhs),
+                                                                   t_values_rhs))
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference);
+}
+TEST_SUITE_END() // GEMMLowpMatrixMultiplyReshapedOnlyRHS
+TEST_SUITE_END() // CL
+} // namespace validation
+} // namespace test
+} // namespace arm_compute
\ No newline at end of file
diff --git a/tests/validation/CL/GEMMMatrixMultiplyReshapedOnlyRHS.cpp b/tests/validation/CL/GEMMMatrixMultiplyReshapedOnlyRHS.cpp
new file mode 100644
index 0000000..83051d2
--- /dev/null
+++ b/tests/validation/CL/GEMMMatrixMultiplyReshapedOnlyRHS.cpp
@@ -0,0 +1,350 @@
+/*
+ * Copyright (c) 2019 ARM Limited.
+ *
+ * SPDX-License-Identifier: MIT
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#include "arm_compute/core/CL/kernels/CLGEMMMatrixMultiplyReshapedOnlyRHSKernel.h"
+#include "arm_compute/core/CL/kernels/CLGEMMReshapeRHSMatrixKernel.h"
+#include "arm_compute/core/Types.h"
+#include "arm_compute/core/utils/misc/ShapeCalculator.h"
+#include "arm_compute/runtime/CL/CLTensor.h"
+#include "arm_compute/runtime/CL/CLTensorAllocator.h"
+#include "tests/CL/CLAccessor.h"
+#include "tests/CL/Helper.h"
+#include "tests/PaddingCalculator.h"
+#include "tests/datasets/ShapeDatasets.h"
+#include "tests/framework/Asserts.h"
+#include "tests/framework/Macros.h"
+#include "tests/framework/datasets/Datasets.h"
+#include "tests/validation/Validation.h"
+#include "tests/validation/fixtures/GEMMFixture.h"
+
+namespace arm_compute
+{
+namespace test
+{
+namespace validation
+{
+using namespace arm_compute::misc::shape_calculator;
+
+// Create function for CLGEMMReshapeRHSMatrixKernel
+using CLGEMMReshapeRHSMatrix = CLSynthetizeFunction<CLGEMMReshapeRHSMatrixKernel>;
+
+// Create function for CLGEMMMatrixMultiplyReshapedOnlyRHSKernel
+using CLGEMMMatrixMultiplyReshapedOnlyRHS = CLSynthetizeFunction<CLGEMMMatrixMultiplyReshapedOnlyRHSKernel>;
+
+// Fixture for CLGEMMMatrixMultiplyReshapedOnlyRHS
+template <typename T>
+using CLGEMMMatrixMultiplyReshapedOnlyRHSFixture = GEMMMatrixMultiplyReshapedOnlyRHSValidationFixture<CLTensor, CLAccessor, T, CLGEMMReshapeRHSMatrix, CLGEMMMatrixMultiplyReshapedOnlyRHS>;
+
+// Fixture for CLGEMMMatrixMultiplyReshapedOnlyRHS3D
+template <typename T>
+using CLGEMMMatrixMultiplyReshapedOnlyRHS3DFixture = GEMMMatrixMultiplyReshapedOnlyRHS3DValidationFixture<CLTensor, CLAccessor, T, CLGEMMReshapeRHSMatrix, CLGEMMMatrixMultiplyReshapedOnlyRHS>;
+
+namespace
+{
+// *INDENT-OFF*
+// clang-format off
+RelativeTolerance<float> rel_tolerance_f32(0.001f);
+constexpr float          abs_tolerance_f32(0.0001f);
+
+RelativeTolerance<half> rel_tolerance_f16(half(0.2));
+constexpr float         tolerance_num_f16 = 0.02f;
+
+/** Alpha values to test - Precommit */
+const auto a_values = framework::dataset::make("alpha", {1.0f, -0.75f} );
+
+/** M values to test */
+const auto m_values = framework::dataset::make("M", 37);
+
+/** M_W values to test */
+const auto m_w_values = framework::dataset::make("M_W", 5);
+
+/** M_H values to test */
+const auto m_h_values = framework::dataset::make("M_H", 7);
+
+/** N values to test */
+const auto n_values = framework::dataset::make("N", 51);
+
+/** K values to test */
+const auto k_values = framework::dataset::make("K", 23);
+
+/** Batch size values to test */
+const auto b_values = framework::dataset::make("batch_size", 1, 3);
+
+/** M0 values to test - Precommit */
+const auto m0_values_precommit = framework::dataset::make("M0", {4, 6});
+
+/** N0 values to test - Precommit */
+const auto n0_values_precommit = framework::dataset::make("N0", { 2, 4 });
+
+/** K0 values to test - Precommit */
+const auto k0_values_precommit = framework::dataset::make("K0", { 4 });
+
+/** H0 values to test - Precommit */
+const auto h0_values_precommit = framework::dataset::make("H0", 1, 3);
+
+/** M0 values to test - Nightly */
+const auto m0_values_nightly = framework::dataset::make("M0", 1, 8);
+
+/** N0 values to test - Nightly */
+const auto n0_values_nightly = framework::dataset::make("N0", { 2, 3, 4, 8 });
+
+/** K0 values to test - Nightly */
+const auto k0_values_nightly = framework::dataset::make("K0", { 2, 3, 4, 8 });
+
+/** H0 values to test - Nightly */
+const auto h0_values_nightly = framework::dataset::make("H0", 1, 4);
+
+/** Interleave values to test with RHS matrix */
+const auto i_values_rhs = framework::dataset::make("interleave_rhs", { true, false });
+
+/** Transpose values to test with RHS matrix */
+const auto t_values_rhs = framework::dataset::make("transpose_rhs", { true, false });
+
+/** Configuration test */
+void validate_configuration(unsigned int m_value, unsigned int n_value, unsigned int k_value, unsigned int b_value, unsigned int m0_value, unsigned int n0_value, unsigned int k0_value, unsigned int h0_value, bool i_value_rhs, bool t_value_rhs, DataType data_type)
+{
+    const unsigned int M = m_value;
+    const unsigned int N = n_value;
+    const unsigned int K = k_value;
+
+    GEMMLHSMatrixInfo lhs_info;
+    lhs_info.m0         = m0_value;
+    lhs_info.k0         = k0_value;
+
+    GEMMRHSMatrixInfo rhs_info;
+    rhs_info.n0         = n0_value;
+    rhs_info.k0         = k0_value;
+    rhs_info.h0         = h0_value;
+    rhs_info.interleave = i_value_rhs;
+    rhs_info.transpose  = t_value_rhs;
+
+    GEMMReshapeInfo gemm_info(M, N, K);
+
+    const TensorShape lhs_shape(K, M, b_value);
+    const TensorShape rhs_shape(N, K, b_value);
+    const TensorShape rhs_shape_reshaped = compute_rhs_reshaped_shape(TensorInfo(rhs_shape, 1, data_type),
+                                                                      rhs_info);
+
+    const TensorShape dst_shape = compute_mm_shape(TensorInfo(lhs_shape, 1, data_type),
+                                                   TensorInfo(rhs_shape_reshaped, 1, data_type),
+                                                   gemm_info);
+
+    // Create tensors
+    CLTensor lhs          = create_tensor<CLTensor>(lhs_shape, data_type);
+    CLTensor rhs_reshaped = create_tensor<CLTensor>(rhs_shape_reshaped, data_type);
+    CLTensor dst          = create_tensor<CLTensor>(dst_shape, data_type);
+
+    ARM_COMPUTE_EXPECT(lhs.info()->is_resizable(), framework::LogLevel::ERRORS);
+    ARM_COMPUTE_EXPECT(rhs_reshaped.info()->is_resizable(), framework::LogLevel::ERRORS);
+    ARM_COMPUTE_EXPECT(dst.info()->is_resizable(), framework::LogLevel::ERRORS);
+
+    // Create and configure function
+    CLGEMMMatrixMultiplyReshapedOnlyRHS gemm;
+    gemm.configure(&lhs, &rhs_reshaped, &dst, 1.0f, lhs_info, rhs_info, gemm_info);
+}
+} // namespace
+
+TEST_SUITE(CL)
+TEST_SUITE(GEMMMatrixMultiplyReshapedOnlyRHS)
+TEST_SUITE(Float)
+TEST_SUITE(FP32)
+DATA_TEST_CASE(Configuration, framework::DatasetMode::ALL, combine(combine(combine(combine(combine(combine(combine(combine(combine(
+                                                                   m_values,
+                                                                   n_values),
+                                                                   k_values),
+                                                                   framework::dataset::make("batch_size", 1)),
+                                                                   m0_values_precommit),
+                                                                   n0_values_precommit),
+                                                                   k0_values_precommit),
+                                                                   h0_values_precommit),
+                                                                   i_values_rhs),
+                                                                   t_values_rhs),
+m_value, n_value, k_value, b_value, m0_value, n0_value, k0_value, h0_value, i_value_rhs, t_value_rhs)
+{
+    validate_configuration(m_value, n_value, k_value, b_value, m0_value, n0_value, k0_value, h0_value, i_value_rhs, t_value_rhs, DataType::F32);
+}
+
+FIXTURE_DATA_TEST_CASE(RunSmall, CLGEMMMatrixMultiplyReshapedOnlyRHSFixture<float>, framework::DatasetMode::ALL,
+                combine(combine(combine(combine(combine(combine(combine(combine(combine(combine(combine(
+                                                                   m_values,
+                                                                   n_values),
+                                                                   k_values),
+                                                                   b_values),
+                                                                   m0_values_precommit),
+                                                                   n0_values_precommit),
+                                                                   k0_values_precommit),
+                                                                   h0_values_precommit),
+                                                                   i_values_rhs),
+                                                                   t_values_rhs),
+                                                                   framework::dataset::make("DataType", DataType::F32)),
+                                                                   a_values))
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference, rel_tolerance_f32, 0.f, abs_tolerance_f32);
+}
+
+FIXTURE_DATA_TEST_CASE(RunLarge, CLGEMMMatrixMultiplyReshapedOnlyRHSFixture<float>, framework::DatasetMode::NIGHTLY,
+                combine(combine(combine(combine(combine(combine(combine(combine(combine(combine(combine(
+                                                                   m_values,
+                                                                   n_values),
+                                                                   k_values),
+                                                                   b_values),
+                                                                   m0_values_nightly),
+                                                                   n0_values_nightly),
+                                                                   k0_values_nightly),
+                                                                   h0_values_nightly),
+                                                                   i_values_rhs),
+                                                                   t_values_rhs),
+                                                                   framework::dataset::make("DataType", DataType::F32)),
+                                                                   a_values))
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference, rel_tolerance_f32, 0.f, abs_tolerance_f32);
+}
+
+FIXTURE_DATA_TEST_CASE(RunSmall3D, CLGEMMMatrixMultiplyReshapedOnlyRHS3DFixture<float>, framework::DatasetMode::ALL,
+                combine(combine(combine(combine(combine(combine(combine(combine(combine(combine(combine(combine(
+                                                                   m_w_values,
+                                                                   m_h_values),
+                                                                   n_values),
+                                                                   k_values),
+                                                                   b_values),
+                                                                   m0_values_precommit),
+                                                                   n0_values_precommit),
+                                                                   k0_values_precommit),
+                                                                   h0_values_precommit),
+                                                                   i_values_rhs),
+                                                                   t_values_rhs),
+                                                                   framework::dataset::make("DataType", DataType::F32)),
+                                                                   a_values))
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference, rel_tolerance_f32, 0.f, abs_tolerance_f32);
+}
+
+FIXTURE_DATA_TEST_CASE(RunLarge3D, CLGEMMMatrixMultiplyReshapedOnlyRHS3DFixture<float>, framework::DatasetMode::NIGHTLY,
+                combine(combine(combine(combine(combine(combine(combine(combine(combine(combine(combine(combine(
+                                                                   m_w_values,
+                                                                   m_h_values),
+                                                                   n_values),
+                                                                   k_values),
+                                                                   b_values),
+                                                                   m0_values_nightly),
+                                                                   n0_values_nightly),
+                                                                   k0_values_nightly),
+                                                                   h0_values_nightly),
+                                                                   i_values_rhs),
+                                                                   t_values_rhs),
+                                                                   framework::dataset::make("DataType", DataType::F32)),
+                                                                   a_values))
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference, rel_tolerance_f32, 0.f, abs_tolerance_f32);
+}
+TEST_SUITE_END() // FP32
+
+TEST_SUITE(FP16)
+FIXTURE_DATA_TEST_CASE(RunSmall, CLGEMMMatrixMultiplyReshapedOnlyRHSFixture<half>, framework::DatasetMode::ALL,
+                combine(combine(combine(combine(combine(combine(combine(combine(combine(combine(combine(
+                                                                   m_values,
+                                                                   n_values),
+                                                                   k_values),
+                                                                   b_values),
+                                                                   m0_values_precommit),
+                                                                   n0_values_precommit),
+                                                                   k0_values_precommit),
+                                                                   h0_values_precommit),
+                                                                   i_values_rhs),
+                                                                   t_values_rhs),
+                                                                   framework::dataset::make("DataType", DataType::F16)),
+                                                                   a_values))
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference, rel_tolerance_f16, tolerance_num_f16);
+}
+
+FIXTURE_DATA_TEST_CASE(RunLarge, CLGEMMMatrixMultiplyReshapedOnlyRHSFixture<half>, framework::DatasetMode::NIGHTLY,
+                combine(combine(combine(combine(combine(combine(combine(combine(combine(combine(combine(
+                                                                   m_values,
+                                                                   n_values),
+                                                                   k_values),
+                                                                   b_values),
+                                                                   m0_values_nightly),
+                                                                   n0_values_nightly),
+                                                                   k0_values_nightly),
+                                                                   h0_values_nightly),
+                                                                   i_values_rhs),
+                                                                   t_values_rhs),
+                                                                   framework::dataset::make("DataType", DataType::F16)),
+                                                                   a_values))
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference, rel_tolerance_f16, tolerance_num_f16);
+}
+
+FIXTURE_DATA_TEST_CASE(RunSmall3D, CLGEMMMatrixMultiplyReshapedOnlyRHS3DFixture<half>, framework::DatasetMode::ALL,
+                combine(combine(combine(combine(combine(combine(combine(combine(combine(combine(combine(combine(
+                                                                   m_w_values,
+                                                                   m_h_values),
+                                                                   n_values),
+                                                                   k_values),
+                                                                   b_values),
+                                                                   m0_values_precommit),
+                                                                   n0_values_precommit),
+                                                                   k0_values_precommit),
+                                                                   h0_values_precommit),
+                                                                   i_values_rhs),
+                                                                   t_values_rhs),
+                                                                   framework::dataset::make("DataType", DataType::F16)),
+                                                                   a_values))
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference, rel_tolerance_f16, tolerance_num_f16);
+}
+
+FIXTURE_DATA_TEST_CASE(RunLarge3D, CLGEMMMatrixMultiplyReshapedOnlyRHS3DFixture<half>, framework::DatasetMode::NIGHTLY,
+                combine(combine(combine(combine(combine(combine(combine(combine(combine(combine(combine(combine(
+                                                                   m_w_values,
+                                                                   m_h_values),
+                                                                   n_values),
+                                                                   k_values),
+                                                                   b_values),
+                                                                   m0_values_nightly),
+                                                                   n0_values_nightly),
+                                                                   k0_values_nightly),
+                                                                   h0_values_nightly),
+                                                                   i_values_rhs),
+                                                                   t_values_rhs),
+                                                                   framework::dataset::make("DataType", DataType::F16)),
+                                                                   a_values))
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference, rel_tolerance_f16, tolerance_num_f16);
+}
+TEST_SUITE_END() // FP16
+TEST_SUITE_END() // Float
+TEST_SUITE_END() // GEMMMatrixMulipltyReshapedOnlyRHS
+TEST_SUITE_END() // CL
+} // namespace validation
+} // namespace test
+} // namespace arm_compute
\ No newline at end of file
diff --git a/tests/validation/CL/HeightConcatenateLayer.cpp b/tests/validation/CL/HeightConcatenateLayer.cpp
new file mode 100644
index 0000000..78182ce
--- /dev/null
+++ b/tests/validation/CL/HeightConcatenateLayer.cpp
@@ -0,0 +1,118 @@
+/*
+ * Copyright (c) 2019 ARM Limited.
+ *
+ * SPDX-License-Identifier: MIT
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#include "arm_compute/core/Types.h"
+#include "arm_compute/runtime/CL/CLTensor.h"
+#include "arm_compute/runtime/CL/CLTensorAllocator.h"
+#include "arm_compute/runtime/CL/functions/CLConcatenateLayer.h"
+#include "tests/CL/CLAccessor.h"
+#include "tests/datasets/ShapeDatasets.h"
+#include "tests/framework/Asserts.h"
+#include "tests/framework/Macros.h"
+#include "tests/framework/datasets/Datasets.h"
+#include "tests/validation/Validation.h"
+
+#include "tests/validation/fixtures/ConcatenateLayerFixture.h"
+
+namespace arm_compute
+{
+namespace test
+{
+namespace validation
+{
+TEST_SUITE(CL)
+TEST_SUITE(HeightConcatenateLayer)
+
+template <typename T>
+using CLHeightConcatenateLayerFixture = ConcatenateLayerValidationFixture<CLTensor, ICLTensor, CLAccessor, CLConcatenateLayer, T>;
+
+TEST_SUITE(Float)
+TEST_SUITE(FP16)
+FIXTURE_DATA_TEST_CASE(RunSmall, CLHeightConcatenateLayerFixture<half>, framework::DatasetMode::PRECOMMIT, combine(combine(concat(datasets::Small2DShapes(), datasets::Tiny4DShapes()),
+                                                                                                                   framework::dataset::make("DataType",
+                                                                                                                           DataType::F16)),
+                                                                                                                   framework::dataset::make("Axis", 1)))
+
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference);
+}
+FIXTURE_DATA_TEST_CASE(RunLarge, CLHeightConcatenateLayerFixture<half>, framework::DatasetMode::NIGHTLY, combine(combine(concat(datasets::Large2DShapes(), datasets::Small4DShapes()),
+                                                                                                                 framework::dataset::make("DataType",
+                                                                                                                         DataType::F16)),
+                                                                                                                 framework::dataset::make("Axis", 1)))
+
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference);
+}
+TEST_SUITE_END()
+
+TEST_SUITE(FP32)
+FIXTURE_DATA_TEST_CASE(RunSmall, CLHeightConcatenateLayerFixture<float>, framework::DatasetMode::PRECOMMIT, combine(combine(concat(datasets::Small2DShapes(), datasets::Tiny4DShapes()),
+                                                                                                                    framework::dataset::make("DataType",
+                                                                                                                            DataType::F32)),
+                                                                                                                    framework::dataset::make("Axis", 1)))
+
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference);
+}
+FIXTURE_DATA_TEST_CASE(RunLarge, CLHeightConcatenateLayerFixture<float>, framework::DatasetMode::NIGHTLY, combine(combine(datasets::ConcatenateLayerShapes(), framework::dataset::make("DataType",
+                                                                                                                  DataType::F32)),
+                                                                                                                  framework::dataset::make("Axis", 1)))
+
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference);
+}
+TEST_SUITE_END()
+TEST_SUITE_END()
+
+TEST_SUITE(Quantized)
+TEST_SUITE(QASYMM8)
+FIXTURE_DATA_TEST_CASE(RunSmall, CLHeightConcatenateLayerFixture<uint8_t>, framework::DatasetMode::PRECOMMIT, combine(combine(concat(datasets::Small2DShapes(), datasets::Tiny4DShapes()),
+                                                                                                                      framework::dataset::make("DataType",
+                                                                                                                              DataType::QASYMM8)),
+                                                                                                                      framework::dataset::make("Axis", 1)))
+
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference);
+}
+FIXTURE_DATA_TEST_CASE(RunLarge, CLHeightConcatenateLayerFixture<uint8_t>, framework::DatasetMode::NIGHTLY, combine(combine(datasets::ConcatenateLayerShapes(), framework::dataset::make("DataType",
+                                                                                                                    DataType::QASYMM8)),
+                                                                                                                    framework::dataset::make("Axis", 1)))
+
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference);
+}
+TEST_SUITE_END()
+TEST_SUITE_END()
+
+TEST_SUITE_END()
+TEST_SUITE_END()
+} // namespace validation
+} // namespace test
+} // namespace arm_compute
diff --git a/tests/validation/CL/LSTMLayer.cpp b/tests/validation/CL/LSTMLayer.cpp
index ea20bd6..71a9383 100644
--- a/tests/validation/CL/LSTMLayer.cpp
+++ b/tests/validation/CL/LSTMLayer.cpp
@@ -109,7 +109,7 @@
                                                         TensorInfo(TensorShape(16U, 2U), 1, DataType::F32),
                                                         TensorInfo(TensorShape(16U, 2U), 1, DataType::F32),
                                                         TensorInfo(TensorShape(16U, 2U), 1, DataType::F32),
-                                                        TensorInfo(TensorShape(11U, 13U), 1, DataType::F32),
+                                                        TensorInfo(TensorShape(11U, 2U), 1, DataType::F32),
                                                         TensorInfo(TensorShape(16U, 2U), 1, DataType::F32),
                })),
                framework::dataset::make("ScratchInfo", { TensorInfo(TensorShape(64U, 2U), 1, DataType::F32),
diff --git a/tests/validation/CL/NonLinearFilter.cpp b/tests/validation/CL/NonLinearFilter.cpp
index 286cc48..536b458 100644
--- a/tests/validation/CL/NonLinearFilter.cpp
+++ b/tests/validation/CL/NonLinearFilter.cpp
@@ -52,8 +52,8 @@
     const uint8_t                          constant_border_value = distribution_u8(generator);
 
     // Create the mask
-    uint8_t mask[mask_size * mask_size];
-    fill_mask_from_pattern(mask, mask_size, mask_size, pattern);
+    std::vector<uint8_t> mask(mask_size * mask_size);
+    fill_mask_from_pattern(mask.data(), mask_size, mask_size, pattern);
     const auto half_mask_size = static_cast<int>(mask_size / 2);
 
     // Create tensors
@@ -65,7 +65,7 @@
 
     // Create and configure function
     CLNonLinearFilter filter;
-    filter.configure(&src, &dst, function, mask_size, pattern, mask, border_mode, constant_border_value);
+    filter.configure(&src, &dst, function, mask_size, pattern, mask.data(), border_mode, constant_border_value);
 
     // Validate valid region
     const ValidRegion dst_valid_region = shape_to_valid_region(shape, border_mode == BorderMode::UNDEFINED, BorderSize(half_mask_size));
diff --git a/tests/validation/CL/PadLayer.cpp b/tests/validation/CL/PadLayer.cpp
index 4bbd7b8..2ad29fc 100644
--- a/tests/validation/CL/PadLayer.cpp
+++ b/tests/validation/CL/PadLayer.cpp
@@ -43,9 +43,9 @@
 const auto PaddingSizesDataset = framework::dataset::make("PaddingSize", { PaddingList{ { 0, 0 } },
     PaddingList{ { 1, 1 } },
     PaddingList{ { 1, 1 }, { 2, 2 } },
-    PaddingList{ { 1, 1 }, { 1, 1 }, { 1, 1 }, { 1, 1 } },
-    PaddingList{ { 0, 0 }, { 1, 0 }, { 0, 1 }, { 1, 2 } },
-    PaddingList{ { 0, 0 }, { 0, 0 }, { 0, 0 }, { 1, 1 } }
+    PaddingList{ { 1, 1 }, { 1, 1 }, { 1, 1 } },
+    PaddingList{ { 0, 0 }, { 1, 0 }, { 0, 1 } },
+    PaddingList{ { 0, 0 }, { 0, 0 }, { 0, 0 } }
 });
 } // namespace
 
@@ -55,32 +55,44 @@
 // *INDENT-OFF*
 // clang-format off
 
-DATA_TEST_CASE(Validate, framework::DatasetMode::ALL, zip(zip(zip(
+DATA_TEST_CASE(Validate, framework::DatasetMode::ALL, zip(zip(zip(zip(
                framework::dataset::make("InputInfo", { TensorInfo(TensorShape(27U, 13U, 2U), 1, DataType::F32),     // Mismatching data type input/output
-                                                       TensorInfo(TensorShape(27U, 13U, 2U), 1, DataType::F32),     // Mismatching shapes
+                                                       TensorInfo(TensorShape(27U, 13U, 2U), 1, DataType::F32),    // Mismatching shapes with padding
                                                        TensorInfo(TensorShape(27U, 13U, 2U), 1, DataType::F32),
+                                                       TensorInfo(TensorShape(27U, 13U, 2U), 1, DataType::F32),    // Mismatching shapes dimension
                                                        TensorInfo(TensorShape(27U, 13U, 2U), 1, DataType::F32),
-                                                       TensorInfo(TensorShape(27U, 13U, 2U), 1, DataType::F32),
-                                                       TensorInfo(TensorShape(32U, 13U, 2U), 1, DataType::F32)
+                                                       TensorInfo(TensorShape(32U, 13U), 1, DataType::F32)     // Invalid padding list
                                                      }),
                framework::dataset::make("OutputInfo",{ TensorInfo(TensorShape(27U, 13U, 2U), 1, DataType::F16),
                                                        TensorInfo(TensorShape(28U, 11U, 2U), 1, DataType::F32),
                                                        TensorInfo(TensorShape(29U, 17U, 2U), 1, DataType::F32),
                                                        TensorInfo(TensorShape(29U, 15U, 4U, 3U), 1, DataType::F32),
-                                                       TensorInfo(TensorShape(27U, 14U, 3U, 4U), 1, DataType::F32),
-                                                       TensorInfo(TensorShape(32U, 13U, 2U, 3U), 1, DataType::F32)
+                                                       TensorInfo(TensorShape(29U, 17U, 2U), 1, DataType::F32),
+                                                       TensorInfo(TensorShape(32U, 13U), 1, DataType::F32)
                                                      })),
                framework::dataset::make("PaddingSize", { PaddingList{{0, 0}},
-                                                      PaddingList{{1, 1}},
-                                                      PaddingList{{1, 1}, {2, 2}},
-                                                      PaddingList{{1,1}, {1,1}, {1,1}, {1,1}},
-                                                      PaddingList{{0,0}, {1,0}, {0,1}, {1,2}},
-                                                      PaddingList{{0,0}, {0,0}, {0,0}, {1,1}}
-                                                      })),
-               framework::dataset::make("Expected", { false, false, true, true, true, true })),
-               input_info, output_info, padding, expected)
+                                                         PaddingList{{1, 1}},
+                                                         PaddingList{{1, 1}, {2, 2}},
+                                                         PaddingList{{1,1}, {1,1}, {1,1}},
+                                                         PaddingList{{1, 1}, {2, 2}},
+                                                         PaddingList{{0,0}, {0,0}, {1,1}}
+                                                         })),
+               framework::dataset::make("PaddingMode", { PaddingMode::CONSTANT,
+                                                         PaddingMode::CONSTANT,
+                                                         PaddingMode::CONSTANT,
+                                                         PaddingMode::SYMMETRIC,
+                                                         PaddingMode::REFLECT,
+                                                         PaddingMode::REFLECT
+})),
+               framework::dataset::make("Expected", { false,
+                                                   false,
+                                                   true,
+                                                   false,
+                                                   true,
+                                                   false })),
+               input_info, output_info, padding, mode, expected)
 {
-    ARM_COMPUTE_EXPECT(bool(CLPadLayer::validate(&input_info.clone()->set_is_resizable(true), &output_info.clone()->set_is_resizable(true), padding)) == expected, framework::LogLevel::ERRORS);
+    ARM_COMPUTE_EXPECT(bool(CLPadLayer::validate(&input_info.clone()->set_is_resizable(true), &output_info.clone()->set_is_resizable(true), padding, PixelValue(), mode)) == expected, framework::LogLevel::ERRORS);
 }
 
 // clang-format on
@@ -92,10 +104,9 @@
 TEST_SUITE(Float)
 
 TEST_SUITE(FP32)
-FIXTURE_DATA_TEST_CASE(RunPadding, CLPaddingFixture<float>, framework::DatasetMode::ALL,
-                       combine(
-                           combine(datasets::SmallShapes(), framework::dataset::make("DataType", { DataType::F32 })),
-                           PaddingSizesDataset))
+FIXTURE_DATA_TEST_CASE(RunSmall, CLPaddingFixture<float>, framework::DatasetMode::ALL,
+                       combine(combine(combine(datasets::Small3DShapes(), framework::dataset::make("DataType", { DataType::F32 })), PaddingSizesDataset),
+                               framework::dataset::make("PaddingMode", { PaddingMode::CONSTANT, PaddingMode::REFLECT, PaddingMode::SYMMETRIC })))
 {
     // Validate output
     validate(CLAccessor(_target), _reference);
@@ -103,10 +114,9 @@
 TEST_SUITE_END() // FP32
 
 TEST_SUITE(FP16)
-FIXTURE_DATA_TEST_CASE(RunPadding, CLPaddingFixture<half>, framework::DatasetMode::ALL,
-                       combine(
-                           combine(datasets::SmallShapes(), framework::dataset::make("DataType", { DataType::F16 })),
-                           PaddingSizesDataset))
+FIXTURE_DATA_TEST_CASE(RunLarge, CLPaddingFixture<half>, framework::DatasetMode::NIGHTLY,
+                       combine(combine(combine(datasets::Large3DShapes(), framework::dataset::make("DataType", { DataType::F16 })), PaddingSizesDataset),
+                               framework::dataset::make("PaddingMode", { PaddingMode::CONSTANT, PaddingMode::REFLECT })))
 {
     // Validate output
     validate(CLAccessor(_target), _reference);
@@ -114,25 +124,18 @@
 TEST_SUITE_END() // FP16
 TEST_SUITE_END() // Float
 
-TEST_SUITE(Integer)
-TEST_SUITE(S8)
-FIXTURE_DATA_TEST_CASE(RunPadding, CLPaddingFixture<int8_t>, framework::DatasetMode::ALL,
-                       combine(
-                           combine(datasets::SmallShapes(), framework::dataset::make("DataType", { DataType::S8 })),
-                           PaddingSizesDataset))
+TEST_SUITE(Quantized)
+TEST_SUITE(QASYMM8)
+FIXTURE_DATA_TEST_CASE(RunSmall, CLPaddingFixture<uint8_t>, framework::DatasetMode::PRECOMMIT,
+                       combine(combine(combine(datasets::Small3DShapes(), framework::dataset::make("DataType", { DataType::QASYMM8 })), PaddingSizesDataset),
+                               framework::dataset::make("PaddingMode", { PaddingMode::CONSTANT, PaddingMode::REFLECT })))
 {
     // Validate output
     validate(CLAccessor(_target), _reference);
 }
-TEST_SUITE_END() // S8
-TEST_SUITE_END() // Integer
-
-TEST_SUITE(Quantized)
-TEST_SUITE(QASYMM8)
-FIXTURE_DATA_TEST_CASE(RunPadding, CLPaddingFixture<uint8_t>, framework::DatasetMode::ALL,
-                       combine(
-                           combine(datasets::SmallShapes(), framework::dataset::make("DataType", { DataType::QASYMM8 })),
-                           PaddingSizesDataset))
+FIXTURE_DATA_TEST_CASE(RunLarge, CLPaddingFixture<uint8_t>, framework::DatasetMode::NIGHTLY,
+                       combine(combine(combine(datasets::Large3DShapes(), framework::dataset::make("DataType", { DataType::QASYMM8 })), PaddingSizesDataset),
+                               framework::dataset::make("PaddingMode", { PaddingMode::CONSTANT, PaddingMode::REFLECT })))
 {
     // Validate output
     validate(CLAccessor(_target), _reference);
diff --git a/tests/validation/CL/PixelWiseMultiplication.cpp b/tests/validation/CL/PixelWiseMultiplication.cpp
index b61ec39..03ce4c9 100644
--- a/tests/validation/CL/PixelWiseMultiplication.cpp
+++ b/tests/validation/CL/PixelWiseMultiplication.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2017-2018 ARM Limited.
+ * Copyright (c) 2017-2019 ARM Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -129,7 +129,7 @@
 FIXTURE_DATA_TEST_CASE(RunSmall, CLPixelWiseMultiplicationQuantizedFixture<uint8_t>, framework::DatasetMode::PRECOMMIT, combine(combine(combine(combine(combine(combine(combine(datasets::SmallShapes(),
                        framework::dataset::make("DataType", DataType::QASYMM8)),
                        framework::dataset::make("Scale", { 1.f, 2.f })),
-                       framework::dataset::make("ConvertPolicy", { ConvertPolicy::SATURATE, ConvertPolicy::WRAP })),
+                       framework::dataset::make("ConvertPolicy", { ConvertPolicy::SATURATE })),
                        framework::dataset::make("RoundingPolicy", RoundingPolicy::TO_NEAREST_EVEN)),
                        framework::dataset::make("QuantizationInfo", { QuantizationInfo(5.f / 255.f, 20) })),
                        framework::dataset::make("QuantizationInfo", { QuantizationInfo(2.f / 255.f, 10) })),
diff --git a/tests/validation/CL/PoolingLayer.cpp b/tests/validation/CL/PoolingLayer.cpp
index 6afafe6..7d79f3f 100644
--- a/tests/validation/CL/PoolingLayer.cpp
+++ b/tests/validation/CL/PoolingLayer.cpp
@@ -74,6 +74,8 @@
 constexpr AbsoluteTolerance<float>   tolerance_f32(0.001f); /**< Tolerance value for comparing reference's output against implementation's output for 32-bit floating-point type */
 constexpr AbsoluteTolerance<float>   tolerance_f16(0.01f);  /**< Tolerance value for comparing reference's output against implementation's output for 16-bit floating-point type */
 constexpr AbsoluteTolerance<uint8_t> tolerance_qasymm8(1);  /**< Tolerance value for comparing reference's output against implementation's output for 8-bit asymmetric type */
+const auto                           pool_data_layout_dataset = framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC });
+
 } // namespace
 
 TEST_SUITE(CL)
@@ -133,7 +135,7 @@
 FIXTURE_DATA_TEST_CASE(RunSmall, CLPoolingLayerFixture<float>, framework::DatasetMode::PRECOMMIT, combine(combine(datasets::SmallShapes(), combine(PoolingLayerDatasetFPSmall,
                                                                                                                   framework::dataset::make("DataType",
                                                                                                                           DataType::F32))),
-                                                                                                          framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })))
+                                                                                                          pool_data_layout_dataset))
 {
     // Validate output
     validate(CLAccessor(_target), _reference, tolerance_f32);
@@ -141,7 +143,7 @@
 FIXTURE_DATA_TEST_CASE(RunLarge, CLPoolingLayerFixture<float>, framework::DatasetMode::NIGHTLY, combine(combine(datasets::LargeShapes(), combine(PoolingLayerDatasetFP,
                                                                                                                 framework::dataset::make("DataType",
                                                                                                                         DataType::F32))),
-                                                                                                        framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })))
+                                                                                                        pool_data_layout_dataset))
 {
     // Validate output
     validate(CLAccessor(_target), _reference, tolerance_f32);
@@ -151,14 +153,14 @@
 TEST_SUITE(FP16)
 FIXTURE_DATA_TEST_CASE(RunSmall, CLPoolingLayerFixture<half>, framework::DatasetMode::PRECOMMIT, combine(combine(datasets::SmallShapes(), combine(PoolingLayerDatasetFPSmall,
                                                                                                                  framework::dataset::make("DataType", DataType::F16))),
-                                                                                                         framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })))
+                                                                                                         pool_data_layout_dataset))
 {
     // Validate output
     validate(CLAccessor(_target), _reference, tolerance_f16);
 }
 FIXTURE_DATA_TEST_CASE(RunLarge, CLPoolingLayerFixture<half>, framework::DatasetMode::NIGHTLY, combine(combine(datasets::LargeShapes(), combine(PoolingLayerDatasetFP,
                                                                                                                framework::dataset::make("DataType", DataType::F16))),
-                                                                                                       framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })))
+                                                                                                       pool_data_layout_dataset))
 {
     // Validate output
     validate(CLAccessor(_target), _reference, tolerance_f16);
@@ -172,20 +174,16 @@
 using CLPoolingLayerQuantizedFixture = PoolingLayerValidationQuantizedFixture<CLTensor, CLAccessor, CLPoolingLayer, T>;
 
 TEST_SUITE(QASYMM8)
-FIXTURE_DATA_TEST_CASE(RunSmall, CLPoolingLayerQuantizedFixture<uint8_t>, framework::DatasetMode::PRECOMMIT, combine(combine(combine(datasets::SmallShapes(), combine(PoolingLayerDatasetQASYMM8Small,
+FIXTURE_DATA_TEST_CASE(RunSmall, CLPoolingLayerQuantizedFixture<uint8_t>, framework::DatasetMode::PRECOMMIT, combine(combine(datasets::SmallShapes(), combine(PoolingLayerDatasetQASYMM8Small,
                                                                                                                      framework::dataset::make("DataType", DataType::QASYMM8))),
-                                                                                                                     framework::dataset::make("QuantizationInfo", { QuantizationInfo(2.f / 255, 127),
-                                                                                                                             QuantizationInfo(7.f / 255, 123)
-                                                                                                                                                                  })),
-                                                                                                                     framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })))
+                                                                                                                     pool_data_layout_dataset))
 {
     // Validate output
     validate(CLAccessor(_target), _reference, tolerance_qasymm8);
 }
-FIXTURE_DATA_TEST_CASE(RunLarge, CLPoolingLayerQuantizedFixture<uint8_t>, framework::DatasetMode::NIGHTLY, combine(combine(combine(datasets::LargeShapes(), combine(PoolingLayerDatasetQASYMM8,
+FIXTURE_DATA_TEST_CASE(RunLarge, CLPoolingLayerQuantizedFixture<uint8_t>, framework::DatasetMode::NIGHTLY, combine(combine(datasets::LargeShapes(), combine(PoolingLayerDatasetQASYMM8,
                                                                                                                    framework::dataset::make("DataType", DataType::QASYMM8))),
-                                                                                                                   framework::dataset::make("QuantizationInfo", { QuantizationInfo(1.f / 255, 0) })),
-                                                                                                                   framework::dataset::make("DataLayout", { DataLayout::NCHW, DataLayout::NHWC })))
+                                                                                                                   pool_data_layout_dataset))
 {
     // Validate output
     validate(CLAccessor(_target), _reference, tolerance_qasymm8);
diff --git a/tests/validation/CL/QuantizationLayer.cpp b/tests/validation/CL/QuantizationLayer.cpp
index f0cc4cc..26e0304 100644
--- a/tests/validation/CL/QuantizationLayer.cpp
+++ b/tests/validation/CL/QuantizationLayer.cpp
@@ -53,21 +53,17 @@
 // *INDENT-OFF*
 // clang-format off
 DATA_TEST_CASE(Validate, framework::DatasetMode::ALL, zip(zip(
-               framework::dataset::make("InputInfo", { TensorInfo(TensorShape(16U, 16U, 16U, 5U), 1, DataType::U8),  // Wrong input data type
-                                                       TensorInfo(TensorShape(16U, 5U, 16U), 1, DataType::U8),       // Invalid shape
+               framework::dataset::make("InputInfo", { TensorInfo(TensorShape(16U, 16U, 16U, 5U), 1, DataType::QASYMM8),  // Wrong input data type
                                                        TensorInfo(TensorShape(16U, 16U, 16U, 5U), 1, DataType::F32), // Wrong output data type
-                                                       TensorInfo(TensorShape(16U, 16U, 2U, 5U), 1, DataType::U8),   // Mismatching shapes
-                                                       TensorInfo(TensorShape(17U, 16U, 16U, 5U), 1, DataType::U8),  // Shrink window
+                                                       TensorInfo(TensorShape(16U, 16U, 2U, 5U), 1, DataType::F32),   // Mismatching shapes
                                                        TensorInfo(TensorShape(16U, 16U, 16U, 5U), 1, DataType::F32), // Valid
                                                      }),
                framework::dataset::make("OutputInfo",{ TensorInfo(TensorShape(16U, 16U, 16U, 5U), 1, DataType::F32),
-                                                       TensorInfo(TensorShape(16U, 5U, 16U), 1, DataType::U8),
                                                        TensorInfo(TensorShape(16U, 16U, 16U, 5U), 1, DataType::U16),
-                                                       TensorInfo(TensorShape(16U, 16U, 16U, 5U), 1, DataType::F32),
-                                                       TensorInfo(TensorShape(17U, 16U, 16U, 5U), 1, DataType::F32),
-                                                       TensorInfo(TensorShape(16U, 16U, 16U, 5U), 1, DataType::U8),
+                                                       TensorInfo(TensorShape(16U, 16U, 16U, 5U), 1, DataType::QASYMM8),
+                                                       TensorInfo(TensorShape(16U, 16U, 16U, 5U), 1, DataType::QASYMM8),
                                                      })),
-               framework::dataset::make("Expected", { false, false, false, false, false, true})),
+               framework::dataset::make("Expected", { false, false, false, true})),
                input_info, output_info, expected)
 {
     ARM_COMPUTE_EXPECT(bool(CLQuantizationLayer::validate(&input_info.clone()->set_is_resizable(false), &output_info.clone()->set_is_resizable(false))) == expected, framework::LogLevel::ERRORS);
@@ -79,7 +75,7 @@
 {
     // Create tensors
     CLTensor src = create_tensor<CLTensor>(shape, data_type);
-    CLTensor dst = create_tensor<CLTensor>(shape, DataType::U8);
+    CLTensor dst = create_tensor<CLTensor>(shape, DataType::QASYMM8);
 
     ARM_COMPUTE_EXPECT(src.info()->is_resizable(), framework::LogLevel::ERRORS);
     ARM_COMPUTE_EXPECT(dst.info()->is_resizable(), framework::LogLevel::ERRORS);
@@ -94,9 +90,8 @@
     validate(dst.info()->valid_region(), valid_region);
 
     // Validate padding
-    const PaddingSize padding = PaddingCalculator(shape.x(), 4).required_padding();
-    validate(src.info()->padding(), padding);
-    validate(dst.info()->padding(), padding);
+    validate(src.info()->padding(), PaddingSize());
+    validate(dst.info()->padding(), PaddingSize());
 }
 
 template <typename T>
@@ -104,19 +99,38 @@
 
 TEST_SUITE(Float)
 TEST_SUITE(FP32)
-FIXTURE_DATA_TEST_CASE(RunSmall, CLQuantizationLayerFixture<float>, framework::DatasetMode::PRECOMMIT, combine(concat(datasets::Small3DShapes(), datasets::Small4DShapes()),
-                                                                                                               framework::dataset::make("DataType", DataType::F32)))
+FIXTURE_DATA_TEST_CASE(RunSmall, CLQuantizationLayerFixture<float>, framework::DatasetMode::PRECOMMIT, combine(combine(concat(datasets::Small3DShapes(), datasets::Small4DShapes()),
+                                                                                                                       framework::dataset::make("DataType", DataType::F32)),
+                                                                                                               framework::dataset::make("QuantizationInfo", { QuantizationInfo(0.5f, 10) })))
 {
     // Validate output
     validate(CLAccessor(_target), _reference, tolerance_f32);
 }
-FIXTURE_DATA_TEST_CASE(RunLarge, CLQuantizationLayerFixture<float>, framework::DatasetMode::NIGHTLY, combine(concat(datasets::Large3DShapes(), datasets::Large4DShapes()),
-                                                                                                             framework::dataset::make("DataType", DataType::F32)))
+FIXTURE_DATA_TEST_CASE(RunLarge, CLQuantizationLayerFixture<float>, framework::DatasetMode::NIGHTLY, combine(combine(concat(datasets::Large3DShapes(), datasets::Large4DShapes()),
+                                                                                                                     framework::dataset::make("DataType", DataType::F32)),
+                                                                                                             framework::dataset::make("QuantizationInfo", { QuantizationInfo(0.5f, 10) })))
 {
     // Validate output
     validate(CLAccessor(_target), _reference, tolerance_f32);
 }
 TEST_SUITE_END() // FP32
+
+TEST_SUITE(FP16)
+FIXTURE_DATA_TEST_CASE(RunSmall, CLQuantizationLayerFixture<half>, framework::DatasetMode::PRECOMMIT, combine(combine(concat(datasets::Small3DShapes(), datasets::Small4DShapes()),
+                                                                                                                      framework::dataset::make("DataType", DataType::F16)),
+                                                                                                              framework::dataset::make("QuantizationInfo", { QuantizationInfo(0.5f, 10) })))
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference, tolerance_f32);
+}
+FIXTURE_DATA_TEST_CASE(RunLarge, CLQuantizationLayerFixture<half>, framework::DatasetMode::NIGHTLY, combine(combine(concat(datasets::Large3DShapes(), datasets::Large4DShapes()),
+                                                                                                                    framework::dataset::make("DataType", DataType::F16)),
+                                                                                                            framework::dataset::make("QuantizationInfo", { QuantizationInfo(0.5f, 10) })))
+{
+    // Validate output
+    validate(CLAccessor(_target), _reference, tolerance_f32);
+}
+TEST_SUITE_END() // FP16
 TEST_SUITE_END() // Float
 
 TEST_SUITE_END() // QuantizationLayer
diff --git a/tests/validation/CL/ReductionOperation.cpp b/tests/validation/CL/ReductionOperation.cpp
index c8474e9..79308c8 100644
--- a/tests/validation/CL/ReductionOperation.cpp
+++ b/tests/validation/CL/ReductionOperation.cpp
@@ -63,7 +63,7 @@
 // clang-format off
 DATA_TEST_CASE(Validate, framework::DatasetMode::ALL, zip(zip(zip(
     framework::dataset::make("InputInfo",          { TensorInfo(TensorShape(128U, 64U), 1, DataType::F32), // Mismatching data type input/output
-                                                     TensorInfo(TensorShape(128U, 64U), 2, DataType::F32), // Number of Input channels != 1
+                                                     TensorInfo(TensorShape(128U, 64U), 3, DataType::F32), // Number of Input channels != 1
                                                      TensorInfo(TensorShape(128U, 64U), 1, DataType::S16), // DataType != QASYMM8/F16/F32
                                                      TensorInfo(TensorShape(128U, 64U), 1, DataType::F32), // Axis >= num_max_dimensions
                                                      TensorInfo(TensorShape(128U, 64U), 1, DataType::QASYMM8), // Axis == 0 and SUM_SQUARE and QASYMM8
diff --git a/tests/validation/CL/Split.cpp b/tests/validation/CL/Split.cpp
index 7736318..e038b89 100644
--- a/tests/validation/CL/Split.cpp
+++ b/tests/validation/CL/Split.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018 ARM Limited.
+ * Copyright (c) 2018-2019 ARM Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -57,6 +57,7 @@
 {
     std::vector<TensorInfo> outputs_info(splits);
     std::vector<ITensorInfo*> outputs_info_ptr;
+    outputs_info_ptr.reserve(splits);
     for(auto &output_info : outputs_info)
     {
         outputs_info_ptr.emplace_back(&output_info);
@@ -76,6 +77,7 @@
     CLTensor                 src = create_tensor<CLTensor>(shape, data_type);
     std::vector<CLTensor>    dsts(splits);
     std::vector<ICLTensor *> dsts_ptrs;
+    dsts_ptrs.reserve(splits);
     for(auto &dst : dsts)
     {
         dsts_ptrs.emplace_back(&dst);
diff --git a/tests/validation/CL/StackLayer.cpp b/tests/validation/CL/StackLayer.cpp
index b62c750..1715a5e 100644
--- a/tests/validation/CL/StackLayer.cpp
+++ b/tests/validation/CL/StackLayer.cpp
@@ -71,13 +71,13 @@
 const auto shapes_1d_large = combine(datasets::Large1DShapes(), framework::dataset::make("Axis", -1, 2));
 
 /** Shapes 2D to test */
-const auto shapes_2d_large = combine(datasets::Large2DShapes(), framework::dataset::make("Axis", -2, 3));
+const auto shapes_2d_large = combine(datasets::Medium2DShapes(), framework::dataset::make("Axis", -2, 3));
 
 /** Shapes 3D to test */
-const auto shapes_3d_large = combine(datasets::Large3DShapes(), framework::dataset::make("Axis", -3, 4));
+const auto shapes_3d_large = combine(datasets::Medium3DShapes(), framework::dataset::make("Axis", -3, 4));
 
 /** Shapes 4D to test */
-const auto shapes_4d_large = combine(datasets::Large4DShapes(), framework::dataset::make("Axis", -4, 5));
+const auto shapes_4d_large = combine(datasets::Medium4DShapes(), framework::dataset::make("Axis", -4, 5));
 } // namespace
 
 /** Fixture to use */
diff --git a/tests/validation/CL/UNIT/TensorAllocator.cpp b/tests/validation/CL/UNIT/TensorAllocator.cpp
index 849eee8..7e47e3d 100644
--- a/tests/validation/CL/UNIT/TensorAllocator.cpp
+++ b/tests/validation/CL/UNIT/TensorAllocator.cpp
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018 ARM Limited.
+ * Copyright (c) 2018-2019 ARM Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -25,10 +25,16 @@
 
 #include "arm_compute/runtime/CL/CLMemoryGroup.h"
 #include "arm_compute/runtime/CL/CLScheduler.h"
+#include "arm_compute/runtime/CL/functions/CLActivationLayer.h"
+#include "tests/CL/CLAccessor.h"
+#include "tests/Globals.h"
 #include "tests/framework/Asserts.h"
 #include "tests/framework/Macros.h"
+#include "tests/validation/Validation.h"
+#include "tests/validation/reference/ActivationLayer.h"
 
 #include <memory>
+#include <random>
 
 namespace arm_compute
 {
@@ -36,14 +42,33 @@
 {
 namespace validation
 {
+namespace
+{
+cl_mem import_malloc_memory_helper(void *ptr, size_t size)
+{
+    const cl_import_properties_arm import_properties[] =
+    {
+        CL_IMPORT_TYPE_ARM,
+        CL_IMPORT_TYPE_HOST_ARM,
+        0
+    };
+
+    cl_int err = CL_SUCCESS;
+    cl_mem buf = clImportMemoryARM(CLKernelLibrary::get().context().get(), CL_MEM_READ_WRITE, import_properties, ptr, size, &err);
+    ARM_COMPUTE_ASSERT(err == CL_SUCCESS);
+
+    return buf;
+}
+} // namespace
+
 TEST_SUITE(CL)
 TEST_SUITE(UNIT)
 TEST_SUITE(TensorAllocator)
 
-TEST_CASE(ImportMemory, framework::DatasetMode::ALL)
+TEST_CASE(ImportMemoryBuffer, framework::DatasetMode::ALL)
 {
     // Init tensor info
-    TensorInfo info(TensorShape(24U, 16U, 3U), 1, DataType::F32);
+    const TensorInfo info(TensorShape(24U, 16U, 3U), 1, DataType::F32);
 
     // Allocate memory buffer
     const size_t total_size = info.total_size();
@@ -62,20 +87,88 @@
     ARM_COMPUTE_EXPECT(!bool(t2.allocator()->import_memory(buf)), framework::LogLevel::ERRORS);
     ARM_COMPUTE_EXPECT(t2.info()->is_resizable(), framework::LogLevel::ERRORS);
 
-    // Positive case : Set raw pointer
-    CLTensor t3;
-    t3.allocator()->init(info);
-    ARM_COMPUTE_EXPECT(bool(t3.allocator()->import_memory(buf)), framework::LogLevel::ERRORS);
-    ARM_COMPUTE_EXPECT(!t3.info()->is_resizable(), framework::LogLevel::ERRORS);
-    ARM_COMPUTE_EXPECT(t3.cl_buffer().get() == buf.get(), framework::LogLevel::ERRORS);
-    t3.allocator()->free();
+    // Negative case : Invalid buffer size
+    CLTensor         t3;
+    const TensorInfo info_neg(TensorShape(32U, 16U, 3U), 1, DataType::F32);
+    t3.allocator()->init(info_neg);
+    ARM_COMPUTE_EXPECT(!bool(t3.allocator()->import_memory(buf)), framework::LogLevel::ERRORS);
     ARM_COMPUTE_EXPECT(t3.info()->is_resizable(), framework::LogLevel::ERRORS);
-    ARM_COMPUTE_EXPECT(t3.cl_buffer().get() != buf.get(), framework::LogLevel::ERRORS);
+
+    // Positive case : Set raw pointer
+    CLTensor t4;
+    t4.allocator()->init(info);
+    ARM_COMPUTE_EXPECT(bool(t4.allocator()->import_memory(buf)), framework::LogLevel::ERRORS);
+    ARM_COMPUTE_EXPECT(!t4.info()->is_resizable(), framework::LogLevel::ERRORS);
+    ARM_COMPUTE_EXPECT(t4.cl_buffer().get() == buf.get(), framework::LogLevel::ERRORS);
+    t4.allocator()->free();
+    ARM_COMPUTE_EXPECT(t4.info()->is_resizable(), framework::LogLevel::ERRORS);
+    ARM_COMPUTE_EXPECT(t4.cl_buffer().get() != buf.get(), framework::LogLevel::ERRORS);
 }
 
-TEST_SUITE_END()
-TEST_SUITE_END()
-TEST_SUITE_END()
+TEST_CASE(ImportMemoryMalloc, framework::DatasetMode::ALL)
+{
+    // Check if import extension is supported
+    if(!device_supports_extension(CLKernelLibrary::get().get_device(), "cl_arm_import_memory_host"))
+    {
+        return;
+    }
+    else
+    {
+        const ActivationLayerInfo act_info(ActivationLayerInfo::ActivationFunction::RELU);
+        const TensorShape         shape     = TensorShape(24U, 16U, 3U);
+        const DataType            data_type = DataType::F32;
+
+        // Create tensor
+        const TensorInfo info(shape, 1, data_type);
+        CLTensor         tensor;
+        tensor.allocator()->init(info);
+
+        // Create and configure activation function
+        CLActivationLayer act_func;
+        act_func.configure(&tensor, nullptr, act_info);
+
+        // Allocate and import tensor
+        const size_t total_size_in_elems = tensor.info()->tensor_shape().total_size();
+        const size_t total_size_in_bytes = tensor.info()->total_size();
+        const size_t alignment           = CLKernelLibrary::get().get_device().getInfo<CL_DEVICE_GLOBAL_MEM_CACHELINE_SIZE>();
+        size_t       space               = total_size_in_bytes + alignment;
+        auto         raw_data            = support::cpp14::make_unique<uint8_t[]>(space);
+
+        void *aligned_ptr = raw_data.get();
+        support::cpp11::align(alignment, total_size_in_bytes, aligned_ptr, space);
+
+        cl::Buffer wrapped_buffer(import_malloc_memory_helper(aligned_ptr, total_size_in_bytes));
+        ARM_COMPUTE_EXPECT(bool(tensor.allocator()->import_memory(wrapped_buffer)), framework::LogLevel::ERRORS);
+        ARM_COMPUTE_EXPECT(!tensor.info()->is_resizable(), framework::LogLevel::ERRORS);
+
+        // Fill tensor
+        std::uniform_real_distribution<float> distribution(-5.f, 5.f);
+        std::mt19937                          gen(library->seed());
+        auto                                 *typed_ptr = reinterpret_cast<float *>(aligned_ptr);
+        for(unsigned int i = 0; i < total_size_in_elems; ++i)
+        {
+            typed_ptr[i] = distribution(gen);
+        }
+
+        // Execute function and sync
+        act_func.run();
+        CLScheduler::get().sync();
+
+        // Validate result by checking that the input has no negative values
+        for(unsigned int i = 0; i < total_size_in_elems; ++i)
+        {
+            ARM_COMPUTE_EXPECT(typed_ptr[i] >= 0, framework::LogLevel::ERRORS);
+        }
+
+        // Release resources
+        tensor.allocator()->free();
+        ARM_COMPUTE_EXPECT(tensor.info()->is_resizable(), framework::LogLevel::ERRORS);
+    }
+}
+
+TEST_SUITE_END() // TensorAllocator
+TEST_SUITE_END() // UNIT
+TEST_SUITE_END() // CL
 } // namespace validation
 } // namespace test
 } // namespace arm_compute
diff --git a/tests/validation/CL/WidthConcatenateLayer.cpp b/tests/validation/CL/WidthConcatenateLayer.cpp
index 647e041..52a4e4c 100644
--- a/tests/validation/CL/WidthConcatenateLayer.cpp
+++ b/tests/validation/CL/WidthConcatenateLayer.cpp
@@ -24,14 +24,14 @@
 #include "arm_compute/core/Types.h"
 #include "arm_compute/runtime/CL/CLTensor.h"
 #include "arm_compute/runtime/CL/CLTensorAllocator.h"
-#include "arm_compute/runtime/CL/functions/CLWidthConcatenateLayer.h"
+#include "arm_compute/runtime/CL/functions/CLConcatenateLayer.h"
 #include "tests/CL/CLAccessor.h"
 #include "tests/datasets/ShapeDatasets.h"
 #include "tests/framework/Asserts.h"
 #include "tests/framework/Macros.h"
 #include "tests/framework/datasets/Datasets.h"
 #include "tests/validation/Validation.h"
-#include "tests/validation/fixtures/WidthConcatenateLayerFixture.h"
+#include "tests/validation/fixtures/ConcatenateLayerFixture.h"
 
 namespace arm_compute
 {
@@ -47,19 +47,24 @@
               framework::dataset::make("InputInfo1", {  TensorInfo(TensorShape(23U, 27U, 5U), 1, DataType::F32), // Mismatching data type input/output
                                                         TensorInfo(TensorShape(23U, 27U, 5U), 1, DataType::F32), // Mismatching y dimension
                                                         TensorInfo(TensorShape(23U, 27U, 5U), 1, DataType::F32), // Mismatching total width
-                                                        TensorInfo(TensorShape(16U, 27U, 5U), 1, DataType::F32)
+                                                        TensorInfo(TensorShape(16U, 27U, 5U), 1, DataType::F32),
+                                                        TensorInfo(TensorShape(21U, 35U, 5U), 1, DataType::F32)
+
               }),
               framework::dataset::make("InputInfo2", {  TensorInfo(TensorShape(24U, 27U, 4U), 1, DataType::F32),
                                                         TensorInfo(TensorShape(52U, 27U, 5U), 1, DataType::F32),
                                                         TensorInfo(TensorShape(52U, 27U, 5U), 1, DataType::F32),
-                                                        TensorInfo(TensorShape(16U, 27U, 5U), 1, DataType::F32)
+                                                        TensorInfo(TensorShape(16U, 27U, 5U), 1, DataType::F32),
+                                                        TensorInfo(TensorShape(10U, 35U, 5U), 1, DataType::F32)
               })),
               framework::dataset::make("OutputInfo", {  TensorInfo(TensorShape(47U, 27U, 5U), 1, DataType::F16),
                                                         TensorInfo(TensorShape(75U, 12U, 5U), 1, DataType::F32),
                                                         TensorInfo(TensorShape(11U, 27U, 5U), 1, DataType::F32),
-                                                        TensorInfo(TensorShape(32U, 27U, 5U), 1, DataType::F32)
+                                                        TensorInfo(TensorShape(32U, 27U, 5U), 1, DataType::F32),
+                                                        TensorInfo(TensorShape(31U, 35U, 5U), 1, DataType::F32)
+
               })),
-              framework::dataset::make("Expected", { false, false, false, true })),
+              framework::dataset::make("Expected", { false, false, false, true, true })),
               input_info1, input_info2, output_info,expected)
 {
     std::vector<TensorInfo> inputs_vector_info;
@@ -67,13 +72,13 @@
     inputs_vector_info.emplace_back(std::move(input_info2));
 
     std::vector<ITensorInfo *> inputs_vector_info_raw;
+    inputs_vector_info_raw.reserve(inputs_vector_info.size());
     for(auto &input : inputs_vector_info)
     {
         inputs_vector_info_raw.emplace_back(&input);
     }
 
-    bool is_valid = bool(CLWidthConcatenateLayer::validate(inputs_vector_info_raw,
-                                                           &output_info.clone()->set_is_resizable(false)));
+    bool is_valid = bool(CLConcatenateLayer::validate(inputs_vector_info_raw,&output_info.clone()->set_is_resizable(true), 0));
     ARM_COMPUTE_EXPECT(is_valid == expected, framework::LogLevel::ERRORS);
 }
 // clang-format on
@@ -93,26 +98,30 @@
     ARM_COMPUTE_EXPECT(dst.info()->is_resizable(), framework::LogLevel::ERRORS);
 
     // Create and configure function
-    CLWidthConcatenateLayer concat_layer;
+    CLConcatenateLayer concat_layer;
 
-    concat_layer.configure({ &src1, &src2, &src3 }, &dst);
+    concat_layer.configure({ &src1, &src2, &src3 }, &dst, 0);
 }
 
 template <typename T>
-using CLWidthConcatenateLayerFixture = WidthConcatenateLayerValidationFixture<CLTensor, ICLTensor, CLAccessor, CLWidthConcatenateLayer, T>;
+using CLWidthConcatenateLayerFixture = ConcatenateLayerValidationFixture<CLTensor, ICLTensor, CLAccessor, CLConcatenateLayer, T>;
 
 TEST_SUITE(Float)
 TEST_SUITE(FP16)
-FIXTURE_DATA_TEST_CASE(RunSmall, CLWidthConcatenateLayerFixture<half>, framework::DatasetMode::PRECOMMIT, combine(concat(datasets::Small2DShapes(), datasets::Tiny4DShapes()),
+FIXTURE_DATA_TEST_CASE(RunSmall, CLWidthConcatenateLayerFixture<half>, framework::DatasetMode::PRECOMMIT, combine(combine(concat(datasets::Small2DShapes(), datasets::Tiny4DShapes()),
                                                                                                                   framework::dataset::make("DataType",
-                                                                                                                          DataType::F16)))
+                                                                                                                          DataType::F16)),
+                                                                                                                  framework::dataset::make("Axis", 0)))
+
 {
     // Validate output
     validate(CLAccessor(_target), _reference);
 }
-FIXTURE_DATA_TEST_CASE(RunLarge, CLWidthConcatenateLayerFixture<half>, framework::DatasetMode::NIGHTLY, combine(concat(datasets::Large2DShapes(), datasets::Small4DShapes()),
-                                                                                                                framework::dataset::make("DataType",
-                                                                                                                        DataType::F16)))
+FIXTURE_DATA_TEST_CASE(RunLarge, CLWidthConcatenateLayerFixture<half>, framework::DatasetMode::NIGHTLY, combine(combine(concat(datasets::Large2DShapes(), datasets::Small4DShapes()),
+                                                                                                                        framework::dataset::make("DataType",
+                                                                                                                                DataType::F16)),
+                                                                                                                framework::dataset::make("Axis", 0)))
+
 {
     // Validate output
     validate(CLAccessor(_target), _reference);
@@ -120,15 +129,18 @@
 TEST_SUITE_END()
 
 TEST_SUITE(FP32)
-FIXTURE_DATA_TEST_CASE(RunSmall, CLWidthConcatenateLayerFixture<float>, framework::DatasetMode::PRECOMMIT, combine(concat(datasets::Small2DShapes(), datasets::Tiny4DShapes()),
+FIXTURE_DATA_TEST_CASE(RunSmall, CLWidthConcatenateLayerFixture<float>, framework::DatasetMode::PRECOMMIT, combine(combine(concat(datasets::Small2DShapes(), datasets::Tiny4DShapes()),
                                                                                                                    framework::dataset::make("DataType",
-                                                                                                                           DataType::F32)))
+                                                                                                                           DataType::F32)),
+                                                                                                                   framework::dataset::make("Axis", 0)))
+
 {
     // Validate output
     validate(CLAccessor(_target), _reference);
 }
-FIXTURE_DATA_TEST_CASE(RunLarge, CLWidthConcatenateLayerFixture<float>, framework::DatasetMode::NIGHTLY, combine(datasets::WidthConcatenateLayerShapes(), framework::dataset::make("DataType",
-                                                                                                                 DataType::F32)))
+FIXTURE_DATA_TEST_CASE(RunLarge, CLWidthConcatenateLayerFixture<float>, framework::DatasetMode::NIGHTLY, combine(combine(datasets::ConcatenateLayerShapes(), framework::dataset::make("DataType",
+                                                                                                                 DataType::F32)),
+                                                                                                                 framework::dataset::make("Axis", 0)))
 {
     // Validate output
     validate(CLAccessor(_target), _reference);
@@ -138,15 +150,17 @@
 
 TEST_SUITE(Quantized)
 TEST_SUITE(QASYMM8)
-FIXTURE_DATA_TEST_CASE(RunSmall, CLWidthConcatenateLayerFixture<uint8_t>, framework::DatasetMode::PRECOMMIT, combine(concat(datasets::Small2DShapes(), datasets::Tiny4DShapes()),
+FIXTURE_DATA_TEST_CASE(RunSmall, CLWidthConcatenateLayerFixture<uint8_t>, framework::DatasetMode::PRECOMMIT, combine(combine(concat(datasets::Small2DShapes(), datasets::Tiny4DShapes()),
                                                                                                                      framework::dataset::make("DataType",
-                                                                                                                             DataType::QASYMM8)))
+                                                                                                                             DataType::QASYMM8)),
+                                                                                                                     framework::dataset::make("Axis", 0)))
 {
     // Validate output
     validate(CLAccessor(_target), _reference);
 }
-FIXTURE_DATA_TEST_CASE(RunLarge, CLWidthConcatenateLayerFixture<uint8_t>, framework::DatasetMode::NIGHTLY, combine(datasets::WidthConcatenateLayerShapes(), framework::dataset::make("DataType",
-                                                                                                                   DataType::QASYMM8)))
+FIXTURE_DATA_TEST_CASE(RunLarge, CLWidthConcatenateLayerFixture<uint8_t>, framework::DatasetMode::NIGHTLY, combine(combine(datasets::ConcatenateLayerShapes(), framework::dataset::make("DataType",
+                                                                                                                   DataType::QASYMM8)),
+                                                                                                                   framework::dataset::make("Axis", 0)))
 {
     // Validate output
     validate(CLAccessor(_target), _reference);
diff --git a/tests/validation/CL/Winograd.cpp b/tests/validation/CL/Winograd.cpp
index e744473..62f0335 100644
--- a/tests/validation/CL/Winograd.cpp
+++ b/tests/validation/CL/Winograd.cpp
@@ -54,7 +54,7 @@
 {
 // *INDENT-OFF*
 // clang-format off
-constexpr AbsoluteTolerance<float> tolerance_f32(0.001f);
+constexpr AbsoluteTolerance<float> tolerance_f32(0.002f);
 const AbsoluteTolerance<half> tolerance_f16(half(0.5f));
 constexpr AbsoluteTolerance<float> tolerance_convolution_layer_f32(0.1f);
 const AbsoluteTolerance<half> tolerance_convolution_layer_f16(half(0.4f));
@@ -81,6 +81,11 @@
                                                     framework::dataset::concat(datasets::SmallWinogradInputTransformDataset4x1_5x1(),
                                                                                datasets::SmallWinogradInputTransformDataset1x4_1x5())))));
 
+const auto SmallWinogradInputTransformDatasetNHWC_FP32 = framework::dataset::concat(SmallWinogradInputTransformDatasetNHWC,
+                                                         framework::dataset::concat(datasets::SmallWinogradInputTransformDataset1x2_1x7(),
+                                                         framework::dataset::concat(datasets::SmallWinogradInputTransformDataset2x1_7x1(),
+                                                                                    datasets::SmallWinogradInputTransformDataset2x2_7x7())));
+
 const auto LargeWinogradInputTransformDatasetNCHW =
            framework::dataset::concat(datasets::LargeWinogradInputTransformDataset2x2_3x3(),
            framework::dataset::concat(datasets::LargeWinogradInputTransformDataset2x1_3x1(),
@@ -98,6 +103,12 @@
            framework::dataset::concat(datasets::LargeWinogradInputTransformDataset4x1_5x1(),
                                       datasets::LargeWinogradInputTransformDataset1x4_1x5())));
 
+const auto LargeWinogradInputTransformDatasetNHWC_FP32 =
+           framework::dataset::concat(LargeWinogradInputTransformDatasetNHWC,
+           framework::dataset::concat(datasets::LargeWinogradInputTransformDataset1x2_1x7(),
+           framework::dataset::concat(datasets::LargeWinogradInputTransformDataset2x1_7x1(),
+                                     (datasets::LargeWinogradInputTransformDataset2x2_7x7()))));
+
 // Filter transform
 const auto SmallWinogradFilterTransformDatasetNCHW =
            framework::dataset::concat(combine(datasets::Small3x3Shapes(), framework::dataset::make("OutputTile", { Size2D(2U, 2U), Size2D(4U, 4U) })),
@@ -107,13 +118,19 @@
            framework::dataset::concat(combine(datasets::Small5x1Shapes(), framework::dataset::make("OutputTile", { Size2D(4U, 1U) })),
                                       combine(datasets::Small1x5Shapes(), framework::dataset::make("OutputTile", { Size2D(1U, 4U) })))))));
 
-const auto SmallWinogradFilterTransformDatasetNHWC =
+const auto SmallWinogradFilterTransformDatasetNHWC_F16 =
            framework::dataset::concat(combine(datasets::Small3x3Shapes(), framework::dataset::make("OutputTile", { Size2D(4U, 4U) })),
            framework::dataset::concat(combine(datasets::Small3x1Shapes(), framework::dataset::make("OutputTile", { Size2D(4U, 1U) })),
            framework::dataset::concat(combine(datasets::Small1x3Shapes(), framework::dataset::make("OutputTile", { Size2D(1U, 4U) })),
            framework::dataset::concat(combine(datasets::Small5x5Shapes(), framework::dataset::make("OutputTile", { Size2D(4U, 4U) })),
            framework::dataset::concat(combine(datasets::Small5x1Shapes(), framework::dataset::make("OutputTile", { Size2D(4U, 1U) })),
-                                      combine(datasets::Small1x5Shapes(), framework::dataset::make("OutputTile", { Size2D(1U, 4U) })))))));
+                                     (combine(datasets::Small1x5Shapes(), framework::dataset::make("OutputTile", { Size2D(1U, 4U) }))))))));
+
+const auto SmallWinogradFilterTransformDatasetNHWC_F32 =
+           framework::dataset::concat(SmallWinogradFilterTransformDatasetNHWC_F16,
+           framework::dataset::concat(combine(datasets::Small7x7Shapes(), framework::dataset::make("OutputTile", { Size2D(2U, 2U) })),
+           framework::dataset::concat(combine(datasets::Small7x1Shapes(), framework::dataset::make("OutputTile", { Size2D(2U, 1U) })),
+                                      combine(datasets::Small1x7Shapes(), framework::dataset::make("OutputTile", { Size2D(1U, 2U) })))));
 
 const auto LargeWinogradFilterTransformDatasetNCHW =
            framework::dataset::concat(combine(datasets::Large3x3Shapes(), framework::dataset::make("OutputTile", { Size2D(2U, 2U), Size2D(4U, 4U) })),
@@ -123,7 +140,7 @@
            framework::dataset::concat(combine(datasets::Large5x1Shapes(), framework::dataset::make("OutputTile", { Size2D(4U, 1U) })),
                                       combine(datasets::Large1x5Shapes(), framework::dataset::make("OutputTile", { Size2D(1U, 4U) })))))));
 
-const auto LargeWinogradFilterTransformDatasetNHWC =
+const auto LargeWinogradFilterTransformDatasetNHWC_F16 =
            framework::dataset::concat(combine(datasets::Large3x3Shapes(), framework::dataset::make("OutputTile", { Size2D(4U, 4U) })),
            framework::dataset::concat(combine(datasets::Large3x1Shapes(), framework::dataset::make("OutputTile", { Size2D(4U, 1U) })),
            framework::dataset::concat(combine(datasets::Large1x3Shapes(), framework::dataset::make("OutputTile", { Size2D(1U, 4U) })),
@@ -131,14 +148,24 @@
            framework::dataset::concat(combine(datasets::Large5x1Shapes(), framework::dataset::make("OutputTile", { Size2D(4U, 1U) })),
                                       combine(datasets::Large1x5Shapes(), framework::dataset::make("OutputTile", { Size2D(1U, 4U) })))))));
 
+const auto LargeWinogradFilterTransformDatasetNHWC_F32 =
+           framework::dataset::concat(LargeWinogradFilterTransformDatasetNHWC_F16,
+           framework::dataset::concat(combine(datasets::Large7x7Shapes(), framework::dataset::make("OutputTile", { Size2D(2U, 2U) })),
+           framework::dataset::concat(combine(datasets::Large7x1Shapes(), framework::dataset::make("OutputTile", { Size2D(2U, 1U) })),
+                                      combine(datasets::Large1x7Shapes(), framework::dataset::make("OutputTile", { Size2D(1U, 2U) })))));
+
 // Output transform
 const auto SmallWinogradOutputTransformDatasetNCHW = datasets::SmallWinogradOutputTransformDatasetNCHW();
 
-const auto SmallWinogradOutputTransformDatasetNHWC = datasets::SmallWinogradOutputTransformDatasetNHWC();
+const auto SmallWinogradOutputTransformDatasetNHWC_F16 = datasets::SmallWinogradOutputTransformDatasetNHWC_F16();
+
+const auto SmallWinogradOutputTransformDatasetNHWC_F32 = datasets::SmallWinogradOutputTransformDatasetNHWC_F32();
 
 const auto LargeWinogradOutputTransformDatasetNCHW = datasets::LargeWinogradOutputTransformDatasetNCHW();
 
-const auto LargeWinogradOutputTransformDatasetNHWC = datasets::LargeWinogradOutputTransformDatasetNHWC();
+const auto LargeWinogradOutputTransformDatasetNHWC_F16 = datasets::LargeWinogradOutputTransformDatasetNHWC_F16();
+
+const auto LargeWinogradOutputTransformDatasetNHWC_F32 = datasets::LargeWinogradOutputTransformDatasetNHWC_F32();
 
 //Activation Functions
 const auto ActivationFunctionsDataset = framework::dataset::make("ActivationInfo",
@@ -252,14 +279,14 @@
 }
 TEST_SUITE_END() // FP16
 TEST_SUITE(FP32)
-FIXTURE_DATA_TEST_CASE(RunSmall, CLWinogradInputTransformFixtureFP32, framework::DatasetMode::PRECOMMIT, combine(combine(SmallWinogradInputTransformDatasetNHWC,
+FIXTURE_DATA_TEST_CASE(RunSmall, CLWinogradInputTransformFixtureFP32, framework::DatasetMode::PRECOMMIT, combine(combine(SmallWinogradInputTransformDatasetNHWC_FP32,
                                                                                                                      framework::dataset::make("DataLayout", { DataLayout::NHWC })),
                                                                                                                      framework::dataset::make("DataType", { DataType::F32 })))
 {
     validate(CLAccessor(_target), _reference, tolerance_f32);
 }
 
-FIXTURE_DATA_TEST_CASE(RunLarge, CLWinogradInputTransformFixtureFP32, framework::DatasetMode::NIGHTLY, combine(combine(LargeWinogradInputTransformDatasetNHWC,
+FIXTURE_DATA_TEST_CASE(RunLarge, CLWinogradInputTransformFixtureFP32, framework::DatasetMode::NIGHTLY, combine(combine(LargeWinogradInputTransformDatasetNHWC_FP32,
                                                                                                                    framework::dataset::make("DataLayout", { DataLayout::NHWC })),
                                                                                                                    framework::dataset::make("DataType", { DataType::F32 })))
 {
@@ -352,7 +379,7 @@
 TEST_SUITE(NHWC)
 TEST_SUITE(FP16)
 FIXTURE_DATA_TEST_CASE(RunSmall, CLWinogradFilterTransformFixtureFP16, framework::DatasetMode::PRECOMMIT,
-                       combine(combine(SmallWinogradFilterTransformDatasetNHWC,
+                       combine(combine(SmallWinogradFilterTransformDatasetNHWC_F16,
                                        framework::dataset::make("DataLayout", { DataLayout::NHWC })),
                                        framework::dataset::make("DataType", { DataType::F16 })))
 {
@@ -361,7 +388,7 @@
 }
 
 FIXTURE_DATA_TEST_CASE(RunLarge, CLWinogradFilterTransformFixtureFP16, framework::DatasetMode::NIGHTLY,
-                       combine(combine(LargeWinogradFilterTransformDatasetNHWC,
+                       combine(combine(LargeWinogradFilterTransformDatasetNHWC_F16,
                                        framework::dataset::make("DataLayout", { DataLayout::NHWC })),
                                        framework::dataset::make("DataType", { DataType::F16 })))
 {
@@ -371,7 +398,7 @@
 TEST_SUITE_END() // FP16
 TEST_SUITE(FP32)
 FIXTURE_DATA_TEST_CASE(RunSmall, CLWinogradFilterTransformFixtureFP32, framework::DatasetMode::PRECOMMIT,
-                       combine(combine(SmallWinogradFilterTransformDatasetNHWC,
+                       combine(combine(SmallWinogradFilterTransformDatasetNHWC_F32,
                                        framework::dataset::make("DataLayout", { DataLayout::NHWC })),
                                        framework::dataset::make("DataType", { DataType::F32 })))
 {
@@ -380,7 +407,7 @@
 }
 
 FIXTURE_DATA_TEST_CASE(RunLarge, CLWinogradFilterTransformFixtureFP32, framework::DatasetMode::NIGHTLY,
-                       combine(combine(LargeWinogradFilterTransformDatasetNHWC,
+                       combine(combine(LargeWinogradFilterTransformDatasetNHWC_F32,
                                        framework::dataset::make("DataLayout", { DataLayout::NHWC })),
                                        framework::dataset::make("DataType", { DataType::F32 })))
 {
@@ -491,7 +518,7 @@
 TEST_SUITE(NHWC)
 TEST_SUITE(FP16)
 FIXTURE_DATA_TEST_CASE(RunSmall, CLWinogradOutputTransformFixtureFP16, framework::DatasetMode::ALL,
-                       combine(combine(SmallWinogradOutputTransformDatasetNHWC,
+                       combine(combine(SmallWinogradOutputTransformDatasetNHWC_F16,
                                framework::dataset::make("DataType", { DataType::F16 })),
                                framework::dataset::make("ActivationInfo",{ ActivationLayerInfo() }) ))
 {
@@ -500,7 +527,7 @@
 }
 
 FIXTURE_DATA_TEST_CASE(RunLarge, CLWinogradOutputTransformFixtureFP16, framework::DatasetMode::NIGHTLY,
-                       combine(combine(LargeWinogradOutputTransformDatasetNHWC,
+                       combine(combine(LargeWinogradOutputTransformDatasetNHWC_F16,
                                framework::dataset::make("DataType", { DataType::F16 })),
                                framework::dataset::make("ActivationInfo",{ ActivationLayerInfo() }) ))
 {
@@ -510,7 +537,7 @@
 TEST_SUITE_END() // FP16
 TEST_SUITE(FP32)
 FIXTURE_DATA_TEST_CASE(RunSmall, CLWinogradOutputTransformFixtureFP32, framework::DatasetMode::ALL,
-                       combine(combine(SmallWinogradOutputTransformDatasetNHWC,
+                       combine(combine(SmallWinogradOutputTransformDatasetNHWC_F32,
                                framework::dataset::make("DataType", { DataType::F32 })),
                                framework::dataset::make("ActivationInfo",{ ActivationLayerInfo() }) ))
 {
@@ -519,7 +546,7 @@
 }
 
 FIXTURE_DATA_TEST_CASE(RunLarge, CLWinogradOutputTransformFixtureFP32, framework::DatasetMode::NIGHTLY,
-                       combine(combine(LargeWinogradOutputTransformDatasetNHWC,
+                       combine(combine(LargeWinogradOutputTransformDatasetNHWC_F32,
                                framework::dataset::make("DataType", { DataType::F32 })),
                                framework::dataset::make("ActivationInfo",{ ActivationLayerInfo() }) ))
 {