Transcribe

Transcribe multiple files from wav to text with timestamps using Azure Speech Recognition API

Based on https://github.com/Azure-Samples/cognitive-services-speech-sdk/tree/master/quickstart/javascript/browser/from-file

Requirements:

• A subscription key for the Speech service. See Try the speech service for free.
• Files prepared in format of wav, 16kHz sample rate & mono

Try it out

• run npm install
• replace SUBSCRIPTION_KEY & SERVICE_REGION with your Azure config
• run node index.js
• example-*.wav.txt output files are generated for default audio files

Transcribe your files

• Put audio files in wav directory
• replace SPEECH_RECOGNITION_LANGUAGE with language your files are in

node index.js

Connects to Azure, runs this script and transcribes files step-by-step.

For 1h long audio file, Azure needs about 20min to finish. Multiple files are transcripted in paralell, so fiveteen 1 hour files would still take about 20minutes. Each recognized sentence is incrementally appended to its newly created source_file.wav.txt file.

Notes:

• On MacOS you can easily convert to 16kHz & mono using iTunes (music.app) (https://support.apple.com/en-us/HT204310)
• txt files with the same name will be overriden on script run
• Pull Requests adding support for proggramatic mp3 -> wav conversion are welcome