Voice categorization for Bob's Burgers by actor and character.
This repo is intended for preparing media, labeling audio, and training a model to recognize actors and their characters in Bob's Burgers.
The audio labeling UI allows you to view audio/video data using multiple lenses and to multi-tag different 'tracks' of data: word-level timestamps, spectrograms, text, etc.
Convert source data to various video, audio, and text formats using utils/.
To run the internal CDN and web server with just foreman start (after gem install foreman if the foreman gem isn't already installed)