Blame - docs/dyn/speech_v1.speech.html - platform/external/python/google-api-python-client

2017-06-06 18:46:08 -0400

[diff] [blame]

100

# process the request.

101

# request.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

102

"enableSeparateRecognitionPerChannel": True or False, # This needs to be set to `true` explicitly and `audio_channel_count` > 1

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

103

# to get each channel recognized separately. The recognition result will

104

# contain a `channel_tag` field to state which channel that result belongs

105

# to. If this is not true, we will only recognize the first channel. The

106

# request is billed cumulatively for all channels recognized:

107

# `audio_channel_count` multiplied by the length of the audio.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

108

"enableAutomaticPunctuation": True or False, # If 'true', adds punctuation to recognition result hypotheses.

109

# This feature is only available in select languages. Setting this for

110

# requests in other languages has no effect at all.

111

# The default 'false' value does not add punctuation to result hypotheses.

112

"maxAlternatives": 42, # Maximum number of recognition hypotheses to be returned.

Sai Cheemalapati

2017-06-06 18:46:08 -0400

[diff] [blame]

113

# Specifically, the maximum number of `SpeechRecognitionAlternative` messages

114

# within each `SpeechRecognitionResult`.

115

# The server may return fewer than `max_alternatives`.

116

# Valid values are `0`-`30`. A value of `0` or `1` will return a maximum of

117

# one. If omitted, will return a maximum of one.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

118

"encoding": "A String", # Encoding of audio data sent in all `RecognitionAudio` messages.

119

# This field is optional for `FLAC` and `WAV` audio files and required

120

# for all other audio formats. For details, see AudioEncoding.

121

"speechContexts": [ # Array of SpeechContext.

122

# A means to provide context to assist the speech recognition. For more

123

# information, see

124

# [speech

125

# adaptation](https://cloud.google.com/speech-to-text/docs/context-strength).

126

{ # Provides "hints" to the speech recognizer to favor specific words and phrases

127

# in the results.

128

"phrases": [ # A list of strings containing words and phrases "hints" so that

129

# the speech recognition is more likely to recognize them. This can be used

130

# to improve the accuracy for specific words and phrases, for example, if

131

# specific commands are typically spoken by the user. This can also be used

132

# to add additional words to the vocabulary of the recognizer. See

133

# [usage limits](https://cloud.google.com/speech-to-text/quotas#content).

134

#

135

# List items can also be set to classes for groups of words that represent

136

# common concepts that occur in natural language. For example, rather than

137

# providing phrase hints for every month of the year, using the $MONTH class

138

# improves the likelihood of correctly transcribing audio that includes

# months.

"A String",

],

},

],

"model": "A String", # Which model to select for the given request. Select the model

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

145

# best suited to your domain to get best results. If a model is not

146

# explicitly specified, then we auto-select a model based on the parameters

147

# in the RecognitionConfig.

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

148

# <table>

149

# <tr>

150

# <td><b>Model</b></td>

151

# <td><b>Description</b></td>

152

# </tr>

153

# <tr>

154

# <td><code>command_and_search</code></td>

155

# <td>Best for short queries such as voice commands or voice search.</td>

156

# </tr>

157

# <tr>

158

# <td><code>phone_call</code></td>

159

# <td>Best for audio that originated from a phone call (typically

160

# recorded at an 8khz sampling rate).</td>

161

# </tr>

162

# <tr>

163

# <td><code>video</code></td>

164

# <td>Best for audio that originated from from video or includes multiple

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

165

# speakers. Ideally the audio is recorded at a 16khz or greater

166

# sampling rate. This is a premium model that costs more than the

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

167

# standard rate.</td>

168

# </tr>

169

# <tr>

170

# <td><code>default</code></td>

171

# <td>Best for audio that is not one of the specific audio models.

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

172

# For example, long-form audio. Ideally the audio is high-fidelity,

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

173

# recorded at a 16khz or greater sampling rate.</td>

174

# </tr>

175

# </table>

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

176

"audioChannelCount": 42, # The number of channels in the input audio data.

177

# ONLY set this for MULTI-CHANNEL recognition.

178

# Valid values for LINEAR16 and FLAC are `1`-`8`.

179

# Valid values for OGG_OPUS are '1'-'254'.

180

# Valid value for MULAW, AMR, AMR_WB and SPEEX_WITH_HEADER_BYTE is only `1`.

181

# If `0` or omitted, defaults to one channel (mono).

182

# Note: We only recognize the first channel by default.

183

# To perform independent recognition on each channel set

184

# `enable_separate_recognition_per_channel` to 'true'.

185

"diarizationConfig": { # Config to enable speaker diarization. # Config to enable speaker diarization and set additional

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

186

# parameters to make diarization better suited for your application.

187

# Note: When this is enabled, we send all the words from the beginning of the

188

# audio for the top alternative in every consecutive STREAMING responses.

189

# This is done in order to improve our speaker tags as our models learn to

190

# identify the speakers in the conversation over time.

191

# For non-streaming requests, the diarization results will be provided only

192

# in the top alternative of the FINAL SpeechRecognitionResult.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

193

"minSpeakerCount": 42, # Minimum number of speakers in the conversation. This range gives you more

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

194

# flexibility by allowing the system to automatically determine the correct

195

# number of speakers. If not set, the default value is 2.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

196

"maxSpeakerCount": 42, # Maximum number of speakers in the conversation. This range gives you more

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

197

# flexibility by allowing the system to automatically determine the correct

198

# number of speakers. If not set, the default value is 6.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

199

"speakerTag": 42, # Output only. Unused.

200

"enableSpeakerDiarization": True or False, # If 'true', enables speaker detection for each recognized word in

201

# the top alternative of the recognition result using a speaker_tag provided

202

# in the WordInfo.

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

203

},

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

204

"enableWordTimeOffsets": True or False, # If `true`, the top result includes a list of words and

205

# the start and end time offsets (timestamps) for those words. If

206

# `false`, no word-level time offset information is returned. The default is

207

# `false`.

208

"languageCode": "A String", # Required. The language of the supplied audio as a

209

# [BCP-47](https://www.rfc-editor.org/rfc/bcp/bcp47.txt) language tag.

210

# Example: "en-US".

211

# See [Language

212

# Support](https://cloud.google.com/speech-to-text/docs/languages) for a list

213

# of the currently supported language codes.

214

"profanityFilter": True or False, # If set to `true`, the server will attempt to filter out

215

# profanities, replacing all but the initial character in each filtered word

216

# with asterisks, e.g. "f***". If set to `false` or omitted, profanities

217

# won't be filtered out.

218

"useEnhanced": True or False, # Set to true to use an enhanced model for speech recognition.

219

# If `use_enhanced` is set to true and the `model` field is not set, then

220

# an appropriate enhanced model is chosen if an enhanced model exists for

221

# the audio.

222

#

223

# If `use_enhanced` is true and an enhanced version of the specified model

224

# does not exist, then the speech is recognized using the standard version

225

# of the specified model.

226

"metadata": { # Description of audio data to be recognized. # Metadata regarding this request.

227

"recordingDeviceName": "A String", # The device used to make the recording. Examples 'Nexus 5X' or

228

# 'Polycom SoundStation IP 6000' or 'POTS' or 'VoIP' or

229

# 'Cardioid Microphone'.

230

"audioTopic": "A String", # Description of the content. Eg. "Recordings of federal supreme court

231

# hearings from 2012".

232

"originalMimeType": "A String", # Mime type of the original audio file. For example `audio/m4a`,

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

233

# `audio/x-alaw-basic`, `audio/mp3`, `audio/3gpp`.

234

# A list of possible audio mime types is maintained at

235

# http://www.iana.org/assignments/media-types/media-types.xhtml#audio

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

236

"microphoneDistance": "A String", # The audio type that most closely describes the audio being recognized.

237

"industryNaicsCodeOfAudio": 42, # The industry vertical to which this speech recognition request most

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

238

# closely applies. This is most indicative of the topics contained

239

# in the audio. Use the 6-digit NAICS code to identify the industry

240

# vertical - see https://www.naics.com/search/.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

241

"originalMediaType": "A String", # The original media the speech was recorded on.

242

"recordingDeviceType": "A String", # The type of device the speech was recorded with.

243

"interactionType": "A String", # The use case most closely describing the audio content to be recognized.

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

244

},

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

245

"sampleRateHertz": 42, # Sample rate in Hertz of the audio data sent in all

246

# `RecognitionAudio` messages. Valid values are: 8000-48000.

247

# 16000 is optimal. For best results, set the sampling rate of the audio

248

# source to 16000 Hz. If that's not possible, use the native sample rate of

249

# the audio source (instead of re-sampling).

250

# This field is optional for FLAC and WAV audio files, but is

251

# required for all other audio formats. For details, see AudioEncoding.

252

},

253

"audio": { # Contains audio data in the encoding specified in the `RecognitionConfig`. # Required. The audio data to be recognized.

254

# Either `content` or `uri` must be supplied. Supplying both or neither

255

# returns google.rpc.Code.INVALID_ARGUMENT. See

256

# [content limits](https://cloud.google.com/speech-to-text/quotas#content).

257

"uri": "A String", # URI that points to a file that contains audio data bytes as specified in

258

# `RecognitionConfig`. The file must not be compressed (for example, gzip).

259

# Currently, only Google Cloud Storage URIs are

260

# supported, which must be specified in the following format:

261

# `gs://bucket_name/object_name` (other URI formats return

262

# google.rpc.Code.INVALID_ARGUMENT). For more information, see

263

# [Request URIs](https://cloud.google.com/storage/docs/reference-uris).

264

"content": "A String", # The audio data bytes encoded as specified in

265

# `RecognitionConfig`. Note: as with all bytes fields, proto buffers use a

266

# pure binary representation, whereas JSON representations use base64.

Sai Cheemalapati

2017-06-06 18:46:08 -0400

[diff] [blame]

},

}

x__xgafv: string, V1 error format.

Allowed values

1 - v1 error format

2 - v2 error format

Returns:

An object of the form:

277

278

{ # This resource represents a long-running operation that is the result of a

279

# network API call.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

280

"done": True or False, # If the value is `false`, it means the operation is still in progress.

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

281

# If `true`, the operation is completed, and either `error` or `response` is

Sai Cheemalapati

2017-06-06 18:46:08 -0400

[diff] [blame]

282

# available.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

283

"response": { # The normal response of the operation in case of success. If the original

Sai Cheemalapati

2017-06-06 18:46:08 -0400

[diff] [blame]

284

# method returns no data on success, such as `Delete`, the response is

285

# `google.protobuf.Empty`. If the original method is standard

286

# `Get`/`Create`/`Update`, the response should be the resource. For other

287

# methods, the response should have the type `XxxResponse`, where `Xxx`

288

# is the original method name. For example, if the original method name

289

# is `TakeSnapshot()`, the inferred response type is

290

# `TakeSnapshotResponse`.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

291

"a_key": "", # Properties of the object. Contains field @type with type URL.

Sai Cheemalapati

2017-06-06 18:46:08 -0400

[diff] [blame]

292

},

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

293

"name": "A String", # The server-assigned name, which is only unique within the same service that

Sai Cheemalapati

2017-06-06 18:46:08 -0400

[diff] [blame]

294

# originally returns it. If you use the default HTTP mapping, the

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

295

# `name` should be a resource name ending with `operations/{unique_id}`.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

296

"error": { # The `Status` type defines a logical error model that is suitable for # The error result of the operation in case of failure or cancellation.

297

# different programming environments, including REST APIs and RPC APIs. It is

298

# used by [gRPC](https://github.com/grpc). Each `Status` message contains

299

# three pieces of data: error code, error message, and error details.

300

#

301

# You can find out more about this error model and how to work with it in the

302

# [API Design Guide](https://cloud.google.com/apis/design/errors).

303

"details": [ # A list of messages that carry the error details. There is a common set of

304

# message types for APIs to use.

305

{

306

"a_key": "", # Properties of the object. Contains field @type with type URL.

307

},

308

],

309

"code": 42, # The status code, which should be an enum value of google.rpc.Code.

310

"message": "A String", # A developer-facing error message, which should be in English. Any

311

# user-facing error message should be localized and sent in the

312

# google.rpc.Status.details field, or localized by the client.

313

},

314

"metadata": { # Service-specific metadata associated with the operation. It typically

315

# contains progress information and common metadata such as create time.

316

# Some services might not provide such metadata. Any method that returns a

317

# long-running operation should document the metadata type, if any.

318

"a_key": "", # Properties of the object. Contains field @type with type URL.

319

},

Sai Cheemalapati

2017-06-06 18:46:08 -0400

[diff] [blame]

}</pre>

</div>

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

324

<code class="details" id="recognize">recognize(body=None, x__xgafv=None)</code>

Sai Cheemalapati

2017-06-06 18:46:08 -0400

[diff] [blame]

325

<pre>Performs synchronous speech recognition: receive results after all audio

326

has been sent and processed.

327

328

Args:

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

329

body: object, The request body.

Sai Cheemalapati

2017-06-06 18:46:08 -0400

[diff] [blame]

330

The object takes the form of:

331

332

{ # The top-level message sent by the client for the `Recognize` method.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

333

"config": { # Provides information to the recognizer that specifies how to process the # Required. Provides information to the recognizer that specifies how to

Sai Cheemalapati

2017-06-06 18:46:08 -0400

[diff] [blame]

334

# process the request.

335

# request.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

336

"enableSeparateRecognitionPerChannel": True or False, # This needs to be set to `true` explicitly and `audio_channel_count` > 1

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

337

# to get each channel recognized separately. The recognition result will

338

# contain a `channel_tag` field to state which channel that result belongs

339

# to. If this is not true, we will only recognize the first channel. The

340

# request is billed cumulatively for all channels recognized:

341

# `audio_channel_count` multiplied by the length of the audio.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

342

"enableAutomaticPunctuation": True or False, # If 'true', adds punctuation to recognition result hypotheses.

343

# This feature is only available in select languages. Setting this for

344

# requests in other languages has no effect at all.

345

# The default 'false' value does not add punctuation to result hypotheses.

346

"maxAlternatives": 42, # Maximum number of recognition hypotheses to be returned.

Sai Cheemalapati

2017-06-06 18:46:08 -0400

[diff] [blame]

347

# Specifically, the maximum number of `SpeechRecognitionAlternative` messages

348

# within each `SpeechRecognitionResult`.

349

# The server may return fewer than `max_alternatives`.

350

# Valid values are `0`-`30`. A value of `0` or `1` will return a maximum of

351

# one. If omitted, will return a maximum of one.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

352

"encoding": "A String", # Encoding of audio data sent in all `RecognitionAudio` messages.

353

# This field is optional for `FLAC` and `WAV` audio files and required

354

# for all other audio formats. For details, see AudioEncoding.

355

"speechContexts": [ # Array of SpeechContext.

356

# A means to provide context to assist the speech recognition. For more

357

# information, see

358

# [speech

359

# adaptation](https://cloud.google.com/speech-to-text/docs/context-strength).

360

{ # Provides "hints" to the speech recognizer to favor specific words and phrases

361

# in the results.

362

"phrases": [ # A list of strings containing words and phrases "hints" so that

363

# the speech recognition is more likely to recognize them. This can be used

364

# to improve the accuracy for specific words and phrases, for example, if

365

# specific commands are typically spoken by the user. This can also be used

366

# to add additional words to the vocabulary of the recognizer. See

367

# [usage limits](https://cloud.google.com/speech-to-text/quotas#content).

368

#

369

# List items can also be set to classes for groups of words that represent

370

# common concepts that occur in natural language. For example, rather than

371

# providing phrase hints for every month of the year, using the $MONTH class

372

# improves the likelihood of correctly transcribing audio that includes

# months.

"A String",

],

},

],

"model": "A String", # Which model to select for the given request. Select the model

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

379

# best suited to your domain to get best results. If a model is not

380

# explicitly specified, then we auto-select a model based on the parameters

381

# in the RecognitionConfig.

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

382

# <table>

383

# <tr>

384

# <td><b>Model</b></td>

385

# <td><b>Description</b></td>

386

# </tr>

387

# <tr>

388

# <td><code>command_and_search</code></td>

389

# <td>Best for short queries such as voice commands or voice search.</td>

390

# </tr>

391

# <tr>

392

# <td><code>phone_call</code></td>

393

# <td>Best for audio that originated from a phone call (typically

394

# recorded at an 8khz sampling rate).</td>

395

# </tr>

396

# <tr>

397

# <td><code>video</code></td>

398

# <td>Best for audio that originated from from video or includes multiple

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

399

# speakers. Ideally the audio is recorded at a 16khz or greater

400

# sampling rate. This is a premium model that costs more than the

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

401

# standard rate.</td>

402

# </tr>

403

# <tr>

404

# <td><code>default</code></td>

405

# <td>Best for audio that is not one of the specific audio models.

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

406

# For example, long-form audio. Ideally the audio is high-fidelity,

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

407

# recorded at a 16khz or greater sampling rate.</td>

408

# </tr>

409

# </table>

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

410

"audioChannelCount": 42, # The number of channels in the input audio data.

411

# ONLY set this for MULTI-CHANNEL recognition.

412

# Valid values for LINEAR16 and FLAC are `1`-`8`.

413

# Valid values for OGG_OPUS are '1'-'254'.

414

# Valid value for MULAW, AMR, AMR_WB and SPEEX_WITH_HEADER_BYTE is only `1`.

415

# If `0` or omitted, defaults to one channel (mono).

416

# Note: We only recognize the first channel by default.

417

# To perform independent recognition on each channel set

418

# `enable_separate_recognition_per_channel` to 'true'.

419

"diarizationConfig": { # Config to enable speaker diarization. # Config to enable speaker diarization and set additional

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

420

# parameters to make diarization better suited for your application.

421

# Note: When this is enabled, we send all the words from the beginning of the

422

# audio for the top alternative in every consecutive STREAMING responses.

423

# This is done in order to improve our speaker tags as our models learn to

424

# identify the speakers in the conversation over time.

425

# For non-streaming requests, the diarization results will be provided only

426

# in the top alternative of the FINAL SpeechRecognitionResult.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

427

"minSpeakerCount": 42, # Minimum number of speakers in the conversation. This range gives you more

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

428

# flexibility by allowing the system to automatically determine the correct

429

# number of speakers. If not set, the default value is 2.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

430

"maxSpeakerCount": 42, # Maximum number of speakers in the conversation. This range gives you more

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

431

# flexibility by allowing the system to automatically determine the correct

432

# number of speakers. If not set, the default value is 6.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

433

"speakerTag": 42, # Output only. Unused.

434

"enableSpeakerDiarization": True or False, # If 'true', enables speaker detection for each recognized word in

435

# the top alternative of the recognition result using a speaker_tag provided

436

# in the WordInfo.

Dan O'Meara

2020-05-01 07:42:23 -0700

[diff] [blame]

437

},

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

438

"enableWordTimeOffsets": True or False, # If `true`, the top result includes a list of words and

439

# the start and end time offsets (timestamps) for those words. If

440

# `false`, no word-level time offset information is returned. The default is

441

# `false`.

442

"languageCode": "A String", # Required. The language of the supplied audio as a

443

# [BCP-47](https://www.rfc-editor.org/rfc/bcp/bcp47.txt) language tag.

444

# Example: "en-US".

445

# See [Language

446

# Support](https://cloud.google.com/speech-to-text/docs/languages) for a list

447

# of the currently supported language codes.

448

"profanityFilter": True or False, # If set to `true`, the server will attempt to filter out

449

# profanities, replacing all but the initial character in each filtered word

450

# with asterisks, e.g. "f***". If set to `false` or omitted, profanities

451

# won't be filtered out.

452

"useEnhanced": True or False, # Set to true to use an enhanced model for speech recognition.

453

# If `use_enhanced` is set to true and the `model` field is not set, then

454

# an appropriate enhanced model is chosen if an enhanced model exists for

455

# the audio.

456

#

457

# If `use_enhanced` is true and an enhanced version of the specified model

458

# does not exist, then the speech is recognized using the standard version

459

# of the specified model.

460

"metadata": { # Description of audio data to be recognized. # Metadata regarding this request.

461

"recordingDeviceName": "A String", # The device used to make the recording. Examples 'Nexus 5X' or

462

# 'Polycom SoundStation IP 6000' or 'POTS' or 'VoIP' or

463

# 'Cardioid Microphone'.

464

"audioTopic": "A String", # Description of the content. Eg. "Recordings of federal supreme court

465

# hearings from 2012".

466

"originalMimeType": "A String", # Mime type of the original audio file. For example `audio/m4a`,

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

467

# `audio/x-alaw-basic`, `audio/mp3`, `audio/3gpp`.

468

# A list of possible audio mime types is maintained at

469

# http://www.iana.org/assignments/media-types/media-types.xhtml#audio

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

470

"microphoneDistance": "A String", # The audio type that most closely describes the audio being recognized.

471

"industryNaicsCodeOfAudio": 42, # The industry vertical to which this speech recognition request most

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

472

# closely applies. This is most indicative of the topics contained

473

# in the audio. Use the 6-digit NAICS code to identify the industry

474

# vertical - see https://www.naics.com/search/.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

475

"originalMediaType": "A String", # The original media the speech was recorded on.

476

"recordingDeviceType": "A String", # The type of device the speech was recorded with.

477

"interactionType": "A String", # The use case most closely describing the audio content to be recognized.

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

478

},

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

479

"sampleRateHertz": 42, # Sample rate in Hertz of the audio data sent in all

480

# `RecognitionAudio` messages. Valid values are: 8000-48000.

481

# 16000 is optimal. For best results, set the sampling rate of the audio

482

# source to 16000 Hz. If that's not possible, use the native sample rate of

483

# the audio source (instead of re-sampling).

484

# This field is optional for FLAC and WAV audio files, but is

485

# required for all other audio formats. For details, see AudioEncoding.

486

},

487

"audio": { # Contains audio data in the encoding specified in the `RecognitionConfig`. # Required. The audio data to be recognized.

488

# Either `content` or `uri` must be supplied. Supplying both or neither

489

# returns google.rpc.Code.INVALID_ARGUMENT. See

490

# [content limits](https://cloud.google.com/speech-to-text/quotas#content).

491

"uri": "A String", # URI that points to a file that contains audio data bytes as specified in

492

# `RecognitionConfig`. The file must not be compressed (for example, gzip).

493

# Currently, only Google Cloud Storage URIs are

494

# supported, which must be specified in the following format:

495

# `gs://bucket_name/object_name` (other URI formats return

496

# google.rpc.Code.INVALID_ARGUMENT). For more information, see

497

# [Request URIs](https://cloud.google.com/storage/docs/reference-uris).

498

"content": "A String", # The audio data bytes encoded as specified in

499

# `RecognitionConfig`. Note: as with all bytes fields, proto buffers use a

500

# pure binary representation, whereas JSON representations use base64.

Sai Cheemalapati

2017-06-06 18:46:08 -0400

[diff] [blame]

},

}

x__xgafv: string, V1 error format.

Allowed values

1 - v1 error format

2 - v2 error format

Returns:

An object of the form:

511

512

{ # The only message returned to the client by the `Recognize` method. It

513

# contains the result as zero or more sequential `SpeechRecognitionResult`

514

# messages.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

515

"results": [ # Sequential list of transcription results corresponding to

Sai Cheemalapati

2017-06-06 18:46:08 -0400

[diff] [blame]

516

# sequential portions of audio.

517

{ # A speech recognition result corresponding to a portion of the audio.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

518

"alternatives": [ # May contain one or more recognition hypotheses (up to the

Sai Cheemalapati

2017-06-06 18:46:08 -0400

[diff] [blame]

519

# maximum specified in `max_alternatives`).

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

520

# These alternatives are ordered in terms of accuracy, with the top (first)

Sai Cheemalapati

2017-06-06 18:46:08 -0400

[diff] [blame]

521

# alternative being the most probable, as ranked by the recognizer.

522

{ # Alternative hypotheses (a.k.a. n-best list).

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

523

"confidence": 3.14, # The confidence estimate between 0.0 and 1.0. A higher number

Sai Cheemalapati

2017-06-06 18:46:08 -0400

[diff] [blame]

524

# indicates an estimated greater likelihood that the recognized words are

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

525

# correct. This field is set only for the top alternative of a non-streaming

526

# result or, of a streaming result where `is_final=true`.

527

# This field is not guaranteed to be accurate and users should not rely on it

528

# to be always provided.

Sai Cheemalapati

2017-06-06 18:46:08 -0400

[diff] [blame]

529

# The default of 0.0 is a sentinel value indicating `confidence` was not set.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

530

"transcript": "A String", # Transcript text representing the words that the user spoke.

531

"words": [ # A list of word-specific information for each recognized word.

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

532

# Note: When `enable_speaker_diarization` is true, you will see all the words

533

# from the beginning of the audio.

534

{ # Word-specific information for recognized words.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

535

"word": "A String", # The word corresponding to this set of information.

536

"speakerTag": 42, # Output only. A distinct integer value is assigned for every speaker within

537

# the audio. This field specifies which one of those speakers was detected to

538

# have spoken this word. Value ranges from '1' to diarization_speaker_count.

539

# speaker_tag is set if enable_speaker_diarization = 'true' and only in the

540

# top alternative.

541

"endTime": "A String", # Time offset relative to the beginning of the audio,

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

542

# and corresponding to the end of the spoken word.

543

# This field is only set if `enable_word_time_offsets=true` and only

544

# in the top hypothesis.

545

# This is an experimental feature and the accuracy of the time offset can

546

# vary.

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

547

"startTime": "A String", # Time offset relative to the beginning of the audio,

Bu Sun Kim

2019-06-14 16:50:42 -0700

[diff] [blame]

548

# and corresponding to the start of the spoken word.

549

# This field is only set if `enable_word_time_offsets=true` and only

550

# in the top hypothesis.

551

# This is an experimental feature and the accuracy of the time offset can

552

# vary.

553

},

554

],

Sai Cheemalapati

2017-06-06 18:46:08 -0400

[diff] [blame]

555

},

556

],

Bu Sun Kim

2020-05-20 12:08:20 -0700

[diff] [blame]

557

"channelTag": 42, # For multi-channel audio, this is the channel number corresponding to the

558

# recognized result for the audio from that channel.

559

# For audio_channel_count = N, its output values can range from '1' to 'N'.

Sai Cheemalapati