I'm playing around with google's cloud speech api. According to its documents I should be able to use an asynchronous request to access a file stored directly in google storage - https://cloud.google.com/speech/docs/async-recognize
I'm basically implementing this code:
const Speech = require('@google-cloud/speech');
const speech = Speech({
projectId: process.env.PROJECTID, credentials:
JSON Key});
const gcsUri = 'gs://my-bucket/audio.raw';
const request = {
encoding: 'LINEAR16',
sampleRateHertz: 16000,
languageCode: 'en-US'
};
speech.startRecognition(gcsUri, request)
.then((results) => {
const operation = results[0];
// Get a Promise represention of the final result of the job
return operation.promise();
})
.then((results) => {
const transcription = results[0];
console.log(`Transcription: ${transcription}`);
})
.catch((err) => {
console.error('ERROR:', err);
});
...but I'm getting the following error --> 'Error: Requested entity was not found.'
I'm guessing this has something to do with access rights, but I can't figure out what to change. I thought since i'm including my project ID and json key file this should be sufficient.
Any help would be much appreciated!
via user3621334
No comments:
Post a Comment