TransWikia.com

Automatic transcription software for phone audio recordings

Software Recommendations Asked on December 11, 2021

Please recommend software to transcribe audio recording of telephone calls. The software needs to perform the transcriptions automatically, and not depend on a human to perform the transcriptions.

The transcriptions do not need to happen in real-time; we can assume the telephone calls are all pre-recorded.

All calls are in the English language. A single person is one of the parties in each of the calls, and that person can provide voice training samples, if needed. The other party varies in each call, and voice training samples are not available for them.

The calls were recorded in compliance with applicable laws, and privacy needs to be maintained; thus online services will not be appropriate.

Here are the requirements:

  1. Android or Windows (any versions)
  2. Just needs to provide a general idea of the conversation. Does not need to be perfect.
  3. Output of sequences of recognized words is sufficient. Does not need to support punctuation.
  4. Closed-source or open-source is fine. Open-source is preferred, but either is acceptable.
  5. Gratis (free of cost) is better, but up to about 40 euros/dollars is fine.
  6. Must not require transmitting the recordings over the internet.

One Answer

Modern speech recognition requires loads of resources and powerful GPU servers for best accuracy. If you want to run on the phone the accuracy will not be sufficient, it will not handle noisy calls, but still it might work for you. You can try Vosk.

Vosk is an open source speech recognition toolkit which supports 9 languages - English, German, French, Spanish, Portuguese, Chinese, Russian, Turkish, Vietnamese. Vosk works offline with small (50 Mb), but accurate model, zero-latency response with streaming API, reconfigurable vocabulary and speaker identification.

You can try Vosk Android Demo for offline recognizer. It might need to train a special model for phone calls but it will work for wideband recordings out of box, so you can test it.

You'd better transfer calls to secure server for processing though. You can try Vosk Server

Answered by Nikolay Shmyrev on December 11, 2021

Add your own answers!

Ask a Question

Get help from others!

© 2024 TransWikia.com. All rights reserved. Sites we Love: PCI Database, UKBizDB, Menu Kuliner, Sharing RPP