audio-embedding topic

Pytorch port of Google Research's VGGish model used for extracting audio features.

harritaylor