Home

Speech-Separations-with-variable-number-of-sources

Public

Audio source separation model with a Whisper ECAPA-TDNN counter and pre‑trained speechbrain/sepformer-libri3mix and speechbrain/sepformer-wsj02mix for speech separation, implemented with SpeechBrain.

Creat2025-04-26T07:01:57
Update2025-08-14T12:05:42
3
Stars
0
Stars Increase

Related projects