WhisperSpeech – An open source text-to-speech system built by inverting Whisper

Re:WhisperSpeech – An open source text-to-speech system buil... by nickmcc on Wednesday January 17, 2024 @09:32PM (#39036831)

I was looking at video on training a custom voice with Piper, following a tutorial at https://www.youtube.com/watch?v=b_we_jma220, and noticed how the datasets required metadata of the text for the source audio files. This training method by Collabora seems to automate that process and only requires an audio file for training.

[ Parent | on HN ]

Re:WhisperSpeech – An open source text-to-speech system buil... by gmerc on Wednesday January 17, 2024 @11:36PM (#39037900)

Whisper solves it, that’s its purpose.

[ Parent | on HN ]

Re:WhisperSpeech – An open source text-to-speech system buil... by jpcl on Thursday January 18, 2024 @06:20AM (#39040472)

Yup, we are using Whisper to transcribe automatically so we can train the model on just speech recordings, without human transcripts.

This works for any language that is well supported by the OpenAI Whisper model.

[ Parent | on HN ]

Re:WhisperSpeech – An open source text-to-speech system buil... by deskamess on Thursday January 18, 2024 @06:46AM (#39040626)

Where can we find the latest OpenAI language model rankings?

[ Parent | on HN ]

Re:WhisperSpeech – An open source text-to-speech system buil... by jpcl on Thursday January 18, 2024 @08:16AM (#39041288)

There is a plot of language performance on their repo: https://github.com/openai/whisper

I am not aware of a multi-lingual leaderboard for speech recognition models.

[ Parent | on HN ]