Skip to content

This is a fork of the OpenAI Whisper ASR model. It is adapted to decoding speech from neural implant data.

License

Notifications You must be signed in to change notification settings

olivershetler/WhisperBrain

 
 

Repository files navigation

WhisperBrain

WhisperBrain is an extension to the OpenAI Whisper model designed to for use with Speech Brain Computer Interfaces.

Setup

Clone the repository and use pip install -e . to create an editable installation.

See WhisperBrainBenchmarkAudio.ipynb for a demonstration on how to use the audio decoder features of the WhisperBrain model.

Go to WhisperBrainDataPrep.ipynb for the script that converts the data from Willett et al. (2023) into an HDF5 file that is used by the dataset objects in this repository. Note that you will need to rename some folders and remove some bad .mat files for the script to run from start to finish.

Go to WhisperBrainTrain.ipynb for an example of how to train the model.

About

This is a fork of the OpenAI Whisper ASR model. It is adapted to decoding speech from neural implant data.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%