| 12345678910111213141516171819202122232425262728293031323334353637383940414243444546474849 |
- type: com.google.api.codegen.samplegen.v1p2.SampleConfigProto
- schema_version: 1.2.0
- samples:
- - region_tag: speech_adaptation_beta
- title: Speech Adaptation (Cloud Storage)
- description: Transcribe a short audio file with speech adaptation.
- rpc: Recognize
- service: google.cloud.speech.v1p1beta1.Speech
- request:
- - field: audio.uri
- value: "gs://cloud-samples-data/speech/brooklyn_bridge.mp3"
- input_parameter: storage_uri
- comment: URI for audio file in Cloud Storage, e.g. gs://[BUCKET]/[FILE]
- - field: config.speech_contexts[0].phrases[0]
- value: "Brooklyn Bridge"
- input_parameter: phrase
- comment: |
- Phrase "hints" help recognize the specified phrases from your audio.
- - field: config.speech_contexts[0].boost
- value: 20.0
- comment: |
- Hint Boost. This value increases the probability that a specific
- phrase will be recognized over other similar sounding phrases.
- The higher the boost, the higher the chance of false positive
- recognition as well. Can accept wide range of positive values.
- Most use cases are best served with values between 0 and 20.
- Using a binary search happroach may help you find the optimal value.
- - field: config.sample_rate_hertz
- value: 44100
- comment: Sample rate in Hertz of the audio data sent
- - field: config.language_code
- value: "en-US"
- comment: The language of the supplied audio
- - field: config.encoding
- value: MP3
- comment: |
- Encoding of audio data sent. This sample sets this explicitly.
- This field is optional for FLAC and WAV audio formats.
- response:
- - loop:
- variable: result
- collection: $resp.results
- body:
- - comment:
- - First alternative is the most probable result
- - define: alternative = result.alternatives[0]
- - print:
- - "Transcript: %s"
- - alternative.transcript
|
