REST Resource: projects.locations.recognizers

Resource: Recognizer

A Recognizer message. Stores recognition configuration and metadata.

JSON representation 
{
  "name": string,
  "uid": string,
  "displayName": string,
  "model": string,
  "languageCodes": [
    string
  ],
  "defaultRecognitionConfig": {
    object (RecognitionConfig)
  },
  "annotations": {
    string: string,
    ...
  },
  "state": enum (State),
  "createTime": string,
  "updateTime": string,
  "deleteTime": string,
  "expireTime": string,
  "etag": string,
  "reconciling": boolean,
  "kmsKeyName": string,
  "kmsKeyVersionName": string
}
Fields
name

string

Output only. Identifier. The resource name of the Recognizer. Format: projects/{project}/locations/{location}/recognizers/{recognizer}.
uid

string

Output only. System-assigned unique identifier for the Recognizer.
displayName

string

User-settable, human-readable name for the Recognizer. Must be 63 characters or less.
model
(deprecated)

string

Optional. This field is now deprecated. Prefer the model field in the RecognitionConfig message.

Which model to use for recognition requests. Select the model best suited to your domain to get best results.

Guidance for choosing which model to use can be found in the Transcription Models Documentation and the models supported in each region can be found in the Table Of Supported Models.
languageCodes[]
(deprecated)

string

Optional. This field is now deprecated. Prefer the languageCodes field in the RecognitionConfig message.

The language of the supplied audio as a BCP-47 language tag.

Supported languages for each model are listed in the Table of Supported Models.

If additional languages are provided, recognition result will contain recognition in the most likely language detected. The recognition result will include the language tag of the language detected in the audio. When you create or update a Recognizer, these values are stored in normalized BCP-47 form. For example, "en-us" is stored as "en-US".
defaultRecognitionConfig

object (RecognitionConfig)

Default configuration to use for requests with this Recognizer. This can be overwritten by inline configuration in the RecognizeRequest.config field.
annotations

map (key: string, value: string)

Allows users to store small amounts of arbitrary data. Both the key and the value must be 63 characters or less each. At most 100 annotations.

An object containing a list of "key": value pairs. Example: { "name": "wrench", "mass": "1.3kg", "count": "3" }.
state

enum (State)

Output only. The Recognizer lifecycle state.
createTime

string (Timestamp format)

Output only. Creation time.

A timestamp in RFC3339 UTC "Zulu" format, with nanosecond resolution and up to nine fractional digits. Examples: "2014-10-02T15:01:23Z" and "2014-10-02T15:01:23.045123456Z".
updateTime

string (Timestamp format)

Output only. The most recent time this Recognizer was modified.

A timestamp in RFC3339 UTC "Zulu" format, with nanosecond resolution and up to nine fractional digits. Examples: "2014-10-02T15:01:23Z" and "2014-10-02T15:01:23.045123456Z".
deleteTime

string (Timestamp format)

Output only. The time at which this Recognizer was requested for deletion.

A timestamp in RFC3339 UTC "Zulu" format, with nanosecond resolution and up to nine fractional digits. Examples: "2014-10-02T15:01:23Z" and "2014-10-02T15:01:23.045123456Z".
expireTime

string (Timestamp format)

Output only. The time at which this Recognizer will be purged.

A timestamp in RFC3339 UTC "Zulu" format, with nanosecond resolution and up to nine fractional digits. Examples: "2014-10-02T15:01:23Z" and "2014-10-02T15:01:23.045123456Z".
etag

string

Output only. This checksum is computed by the server based on the value of other fields. This may be sent on update, undelete, and delete requests to ensure the client has an up-to-date value before proceeding.
reconciling

boolean

Output only. Whether or not this Recognizer is in the process of being updated.
kmsKeyName

string

Output only. The KMS key name with which the Recognizer is encrypted. The expected format is projects/{project}/locations/{location}/keyRings/{key_ring}/cryptoKeys/{crypto_key}.
kmsKeyVersionName

string

Output only. The KMS key version name with which the Recognizer is encrypted. The expected format is projects/{project}/locations/{location}/keyRings/{key_ring}/cryptoKeys/{crypto_key}/cryptoKeyVersions/{crypto_key_version}.

RecognitionConfig

Provides information to the Recognizer that specifies how to process the recognition request.

JSON representation 
{
  "model": string,
  "languageCodes": [
    string
  ],
  "features": {
    object (RecognitionFeatures)
  },
  "adaptation": {
    object (SpeechAdaptation)
  },
  "transcriptNormalization": {
    object (TranscriptNormalization)
  },
  "translationConfig": {
    object (TranslationConfig)
  },

  // Union field decoding_config can be only one of the following:
  "autoDecodingConfig": {
    object (AutoDetectDecodingConfig)
  },
  "explicitDecodingConfig": {
    object (ExplicitDecodingConfig)
  }
  // End of list of possible types for union field decoding_config.
}
Fields
model

string

Optional. Which model to use for recognition requests. Select the model best suited to your domain to get best results.

Guidance for choosing which model to use can be found in the Transcription Models Documentation and the models supported in each region can be found in the Table Of Supported Models.
languageCodes[]

string

Optional. The language of the supplied audio as a BCP-47 language tag. Language tags are normalized to BCP-47 before they are used eg "en-us" becomes "en-US".

Supported languages for each model are listed in the Table of Supported Models.

If additional languages are provided, recognition result will contain recognition in the most likely language detected. The recognition result will include the language tag of the language detected in the audio.
features

object (RecognitionFeatures)

Speech recognition features to enable.
adaptation

object (SpeechAdaptation)

Speech adaptation context that weights recognizer predictions for specific words and phrases.
transcriptNormalization

object (TranscriptNormalization)

Optional. Use transcription normalization to automatically replace parts of the transcript with phrases of your choosing. For StreamingRecognize, this normalization only applies to stable partial transcripts (stability > 0.8) and final transcripts.
translationConfig

object (TranslationConfig)

Optional. Optional configuration used to automatically run translation on the given audio to the desired language for supported models.
Union field decoding_config. Decoding parameters for audio being sent for recognition. decoding_config can be only one of the following:
autoDecodingConfig

object (AutoDetectDecodingConfig)

Automatically detect decoding parameters. Preferred for supported formats.
explicitDecodingConfig

object (ExplicitDecodingConfig)

Explicitly specified decoding parameters. Required if using headerless PCM audio (linear16, mulaw, alaw).

AutoDetectDecodingConfig

This type has no fields.

Automatically detected decoding parameters. Supported for the following encodings:

  • WAV_LINEAR16: 16-bit signed little-endian PCM samples in a WAV container.

  • WAV_MULAW: 8-bit companded mulaw samples in a WAV container.

  • WAV_ALAW: 8-bit companded alaw samples in a WAV container.

  • RFC4867_5_AMR: AMR frames with an rfc4867.5 header.

  • RFC4867_5_AMRWB: AMR-WB frames with an rfc4867.5 header.

  • FLAC: FLAC frames in the "native FLAC" container format.

  • MP3: MPEG audio frames with optional (ignored) ID3 metadata.

  • OGG_OPUS: Opus audio frames in an Ogg container.

  • WEBM_OPUS: Opus audio frames in a WebM container.

  • MP4_AAC: AAC audio frames in an MP4 container.

  • M4A_AAC: AAC audio frames in an M4A container.

  • MOV_AAC: AAC audio frames in an MOV container.

ExplicitDecodingConfig

Explicitly specified decoding parameters.

JSON representation 
{
  "encoding": enum (AudioEncoding),
  "sampleRateHertz": integer,
  "audioChannelCount": integer
}
Fields
encoding

enum (AudioEncoding)

Required. Encoding of the audio data sent for recognition.
sampleRateHertz

integer

Optional. Sample rate in Hertz of the audio data sent for recognition. Valid values are: 8000-48000, and 16000 is optimal. For best results, set the sampling rate of the audio source to 16000 Hz. If that's not possible, use the native sample rate of the audio source (instead of resampling). Note that this field is marked as OPTIONAL for backward compatibility reasons. It is (and has always been) effectively REQUIRED.
audioChannelCount

integer

Optional. Number of channels present in the audio data sent for recognition. Note that this field is marked as OPTIONAL for backward compatibility reasons. It is (and has always been) effectively REQUIRED.

The maximum allowed value is 8.

AudioEncoding

Supported audio data encodings.

Enums
AUDIO_ENCODING_UNSPECIFIED Default value. This value is unused.
LINEAR16 Headerless 16-bit signed little-endian PCM samples.
MULAW Headerless 8-bit companded mulaw samples.
ALAW Headerless 8-bit companded alaw samples.
AMR AMR frames with an rfc4867.5 header.
AMR_WB AMR-WB frames with an rfc4867.5 header.
FLAC FLAC frames in the "native FLAC" container format.
MP3 MPEG audio frames with optional (ignored) ID3 metadata.
OGG_OPUS Opus audio frames in an Ogg container.
WEBM_OPUS Opus audio frames in a WebM container.
MP4_AAC AAC audio frames in an MP4 container.
M4A_AAC AAC audio frames in an M4A container.
MOV_AAC AAC audio frames in an MOV container.

RecognitionFeatures

Available recognition features.

JSON representation 
{
  "profanityFilter": boolean,
  "enableWordTimeOffsets": boolean,
  "enableWordConfidence": boolean,
  "enableAutomaticPunctuation": boolean,
  "enableSpokenPunctuation": boolean,
  "enableSpokenEmojis": boolean,
  "multiChannelMode": enum (MultiChannelMode),
  "diarizationConfig": {
    object (SpeakerDiarizationConfig)
  },
  "maxAlternatives": integer
}
Fields
profanityFilter

boolean

If set to true, the server will attempt to filter out profanities, replacing all but the initial character in each filtered word with asterisks, for instance, "f***". If set to false or omitted, profanities won't be filtered out.
enableWordTimeOffsets

boolean

If true, the top result includes a list of words and the start and end time offsets (timestamps) for those words. If false, no word-level time offset information is returned. The default is false.
enableWordConfidence

boolean

If true, the top result includes a list of words and the confidence for those words. If false, no word-level confidence information is returned. The default is false.
enableAutomaticPunctuation

boolean

If true, adds punctuation to recognition result hypotheses. This feature is only available in select languages. The default false value does not add punctuation to result hypotheses.
enableSpokenPunctuation

boolean

The spoken punctuation behavior for the call. If true, replaces spoken punctuation with the corresponding symbols in the request. For example, "how are you question mark" becomes "how are you?". See https://cloud.google.com/speech-to-text/docs/spoken-punctuation for support. If false, spoken punctuation is not replaced.
enableSpokenEmojis

boolean

The spoken emoji behavior for the call. If true, adds spoken emoji formatting for the request. This will replace spoken emojis with the corresponding Unicode symbols in the final transcript. If false, spoken emojis are not replaced.
multiChannelMode

enum (MultiChannelMode)

Mode for recognizing multi-channel audio.
diarizationConfig

object (SpeakerDiarizationConfig)

Configuration to enable speaker diarization and set additional parameters to make diarization better suited for your application. When this is enabled, we send all the words from the beginning of the audio for the top alternative in every consecutive STREAMING responses. This is done in order to improve our speaker tags as our models learn to identify the speakers in the conversation over time. For non-streaming requests, the diarization results will be provided only in the top alternative of the FINAL SpeechRecognitionResult.
maxAlternatives

integer

Maximum number of recognition hypotheses to be returned. The server may return fewer than maxAlternatives. Valid values are 0-30. A value of 0 or 1 will return a maximum of one. If omitted, will return a maximum of one.

MultiChannelMode

Options for how to recognize multi-channel audio.

Enums
MULTI_CHANNEL_MODE_UNSPECIFIED Default value for the multi-channel mode. If the audio contains multiple channels, only the first channel will be transcribed; other channels will be ignored.
SEPARATE_RECOGNITION_PER_CHANNEL If selected, each channel in the provided audio is transcribed independently. This cannot be selected if the selected model is latest_short.

SpeakerDiarizationConfig

Configuration to enable speaker diarization.

JSON representation 
{
  "minSpeakerCount": integer,
  "maxSpeakerCount": integer
}
Fields
minSpeakerCount

integer

Required. Minimum number of speakers in the conversation. This range gives you more flexibility by allowing the system to automatically determine the correct number of speakers.

To fix the number of speakers detected in the audio, set minSpeakerCount = maxSpeakerCount.
maxSpeakerCount

integer

Required. Maximum number of speakers in the conversation. Valid values are: 1-6. Must be >= minSpeakerCount. This range gives you more flexibility by allowing the system to automatically determine the correct number of speakers.

SpeechAdaptation

Provides "hints" to the speech recognizer to favor specific words and phrases in the results. PhraseSets can be specified as an inline resource, or a reference to an existing PhraseSet resource.

JSON representation 
{
  "phraseSets": [
    {
      object (AdaptationPhraseSet)
    }
  ],
  "customClasses": [
    {
      object (CustomClass)
    }
  ]
}
Fields
phraseSets[]

object (AdaptationPhraseSet)

A list of inline or referenced PhraseSets.
customClasses[]

object (CustomClass)

A list of inline CustomClasses. Existing CustomClass resources can be referenced directly in a PhraseSet.

AdaptationPhraseSet

A biasing PhraseSet, which can be either a string referencing the name of an existing PhraseSets resource, or an inline definition of a PhraseSet.

JSON representation 
{

  // Union field value can be only one of the following:
  "phraseSet": string,
  "inlinePhraseSet": {
    object (PhraseSet)
  }
  // End of list of possible types for union field value.
}
Fields

Union field value.

value can be only one of the following:
phraseSet

string

The name of an existing PhraseSet resource. The user must have read access to the resource and it must not be deleted.
inlinePhraseSet

object (PhraseSet)

An inline defined PhraseSet.

TranscriptNormalization

Transcription normalization configuration. Use transcription normalization to automatically replace parts of the transcript with phrases of your choosing. For StreamingRecognize, this normalization only applies to stable partial transcripts (stability > 0.8) and final transcripts.

JSON representation 
{
  "entries": [
    {
      object (Entry)
    }
  ]
}
Fields
entries[]

object (Entry)

A list of replacement entries. We will perform replacement with one entry at a time. For example, the second entry in ["cat" => "dog", "mountain cat" => "mountain dog"] will never be applied because we will always process the first entry before it. At most 100 entries.

Entry

A single replacement configuration.

JSON representation 
{
  "search": string,
  "replace": string,
  "caseSensitive": boolean
}
Fields
search

string

What to replace. Max length is 100 characters.
replace

string

What to replace with. Max length is 100 characters.
caseSensitive

boolean

Whether the search is case sensitive.

TranslationConfig

Translation configuration. Use to translate the given audio into text for the desired language.

JSON representation 
{
  "targetLanguage": string
}
Fields
targetLanguage

string

Required. The language code to translate to.

State

Set of states that define the lifecycle of a Recognizer.

Enums
STATE_UNSPECIFIED The default value. This value is used if the state is omitted.
ACTIVE The Recognizer is active and ready for use.
DELETED This Recognizer has been deleted.

Methods

batchRecognize

 Performs batch asynchronous speech recognition: send a request with N audio files and receive a long running operation that can be polled to see when the transcriptions are finished.

create

 Creates a Recognizer.

delete

 Deletes the Recognizer.

get

 Returns the requested Recognizer.

list

 Lists Recognizers.

patch

 Updates the Recognizer.

recognize

 Performs synchronous Speech recognition: receive results after all audio has been sent and processed.

undelete

 Undeletes the Recognizer.