Speech to text conversion
speech/recognize@1.0.04 providers
Speech recognition
Real-time speech recognition.
Input
Audio content
Language code
Audio encoding
Maximum alternatives
Result
Results
1.Choose a provider
2.Use Recognize with mock in your code
Below instructions are for our Node.js SDK. Use OneService for other languages.
npm i @superfaceai/one-sdk
Copied!
const { SuperfaceClient } = require('@superfaceai/one-sdk');
// You can manage tokens here: https://superface.ai/insights
const sdk = new SuperfaceClient({ sdkAuthToken: '<< Login to get your token >>' });
async function run() {
// Load the profile
const profile = await sdk.getProfile('speech/recognize@1.0.0');
// Use the profile
const result = await profile
.getUseCase('Recognize')
.perform({
audioContent: '<base64 encoded wav audio>',
languageCode: 'en-US'
}, {
provider: 'mock'
});
// Handle the result
try {
const data = result.unwrap();
console.log(data);
} catch (error) {
console.error(error);
}
}
run();Copied!
Structure details
Input (object)
audioContent
required
Audio data in the encoding specified by audioEncodig input parameter.
languageCode
required
The language (and potentially also the region) of the speech expressed as a BCP-47 language tag, e.g. 'en-US'.
string
audioEncoding
optional
Encoding of audio data sent. This input is optional for WAV audio files and required for other audio formats.
enum
wav
maxAlternatives
optional
Maximum number of recognition hypotheses to be returned. The server may return fewer than maxAlternatives. Valid values are 0-30. Default value is 1.
number
Example
{
"audioContent": "<base64 encoded wav audio>",
"languageCode": "en-US"
}Copied!
Result (object)
results
required
Sequential list of transcription results corresponding to sequential portions of audio.
list
alternatives
required
Alternative hypotheses.
list
transcript
required
Transcript text representing the words recognized in audio input.
string
confidence
optional
The confidence estimate between 0.0 and 1.0. A higher number indicates an estimated greater likelihood that the recognized words are correct.
The default of 0.0 is a sentinel value indicating confidence was not set.
number
Example
{
"results": [
{
"alternatives": [
{
"confidence": 0.8393012,
"transcript": "hello world"
}
]
}
]
}Copied!