package asr

Get desktop application:
View/edit binary Protocol Buffers messages

• optional CloudSpeechStreamObserverParams observer_params = 1

  Parameters for the observer, who converts the speech protos into TranscriptionResult.

• optional bool filter_profanity = 2

  Allows profanity to be filtered by the Cloud Speech API.

• optional CloudSpeechSessionParams.EncoderParams encoder_params = 3

Details about how the audio signal should be compressed prior to sending it to the server.

Used in: CloudSpeechSessionParams

• optional bool enable_encoder = 1

  If the encoder isn't supported, uncompressed audio will be used. When this is false, other EncoderParams fields are ignored.

• optional audio.CodecAndBitrate codec = 2

• optional bool allow_vbr = 3

  Uses variable bitrate encoding, if available. Currently this is available for OggOpus only.

Used in: CloudSpeechSessionParams

• optional bool reject_unstable_hypotheses = 1

  When true, hypotheses are held a bit longer until they likely will not change again before being displayed.

• optional string locale = 1

  Required.

• optional SpeechRecognitionModelOptions.SpecificModel model = 2

• repeated string bias_words = 3

  Words to be passed to the speech recognizer as bias. It is up to each implementation to decide whether these will be used or not.

Select which model to use. Not all models are necessarily available for all recognition systems or locales. It is up to the individual session to warn the user about availability

Used in: SpeechRecognitionModelOptions

• DICTATION_DEFAULT = 0

• VIDEO = 1

• optional string text = 1

  An utterance level copy of the text.

• optional float confidence = 2

  Confidence for the whole utterance [0, 1].

• optional google.protobuf.Timestamp start_timestamp = 7

  The epoch time at which the utterance was started.

• optional google.protobuf.Timestamp end_timestamp = 3

  The epoch time at which the utterance was completed.

• optional audio.SpeakerIdInfo speaker_info = 5

  The identity of the speaker.

• repeated TranscriptionResult.Word word_level_detail = 4

  Word-level detail. NOTE: Some recognizers (namely the CloudSpeech API) do not give fine-grain information until results are finalized.

• optional string language_code = 6

  The language code in this result. See https://cloud.google.com/speech-to-text/docs/languages for more details. For example, English (United States) : en-US Chinese, Mandarin (Traditional, Taiwan) : cmn-Hant-TW

Fine-grain information about each word. NOTE: the TranscriptResultFormatter may colorize the coarse-grain transcript by the corresponding word information such as confidence and speaker_id if fine-grain word_level_detail is not empty.

Used in: TranscriptionResult

• optional string text = 1

• optional float confidence = 2

  Confidence for just this word [0, 1].

• optional audio.SpeakerIdInfo speaker_info = 3

  An integer tag for the identity of the active speaker.

• optional google.protobuf.Timestamp start_timestamp = 4

  The time at which the word was started.

• optional google.protobuf.Timestamp end_timestamp = 5

  The time at which the word was completed.

• optional google.protobuf.Duration extended_silence_duration_for_line_breaks = 1

  Silences longer than this will cause a space to be inserted.

• optional int32 num_extended_silence_line_breaks = 2

  Number of '\n' characters to add in the event of extended silence. 1 moves to the next line, 2 leaves a blank space in between two lines, and so on...

• optional int32 num_language_switch_line_breaks = 3

  Number of '\n' characters to add in the event of language switch. 1 moves to the next line, 2 leaves a blank space in between two lines, and so on...

• optional bool italicize_current_hypothesis = 4

  Put current hypotheses in italics.

• optional TranscriptionResultFormatterOptions.TranscriptColoringStyle transcript_coloring_style = 5

  If true, use a yellow->blue colormap to indicate confidence.

• optional TranscriptionResultFormatterOptions.TextColormap text_colormap = 6

  The color theme used for the text.

• optional TranscriptionResultFormatterOptions.SpeakerIndicationStyle speaker_indication_style = 7

  A label that indicates which speaker is active.

Used in: TranscriptionResultFormatterOptions

• UNSPECIFIED_SPEAKER_INDICATION_STYLE = 0

• NO_SPEAKER_INDICATION = 1

• SHOW_SPEAKER_NUMBER = 2

Color selection for the text (does not change background). Dark colors for a black-on-white theme. Bright colors for a white-on-black theme.

Used in: TranscriptionResultFormatterOptions

• UNSPECIFIED_THEME = 0

• LIGHT_THEME = 1

• DARK_THEME = 2

Details on the manner in which the transcript will be colored.

Used in: TranscriptionResultFormatterOptions

• UNSPECIFIED_COLORING_STYLE = 0

  Will do NO_COLORING.

• NO_COLORING = 1

• COLOR_BY_UTTERANCE_LEVEL_CONFIDENCE = 2

• COLOR_BY_WORD_LEVEL_CONFIDENCE = 3

• COLOR_BY_SPEAKER_ID = 4