asteroid-team/asteroid
The PyTorch-based audio source separation toolkit for researchers
Provides modular architectures (ConvTasNet, DPRNN, DPTNet, etc.) with permutation invariant training (PIT) for flexible model composition, and includes curated recipes reproducing published papers on standard benchmarks (WSJ0-mix, WHAM, LibriMix, MUSDB18). Integrates with TensorBoard for experiment tracking and supports diverse datasets from speech separation to music source separation tasks.
2,547 stars and 39,702 monthly downloads. No commits in the last 6 months. Available on PyPI.
Stars
2,547
Forks
446
Language
Python
License
MIT
Category
Last pushed
Oct 06, 2025
Monthly downloads
39,702
Commits (30d)
0
Dependencies
15
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/asteroid-team/asteroid"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related frameworks
pytorch/audio
Data manipulation and transformation for audio signal processing, powered by PyTorch
deezer/spleeter
Deezer source separation library including pretrained models.
audeering/opensmile
The Munich Open-Source Large-Scale Multimedia Feature Extractor
drscotthawley/aeiou
(ML) audio engineering i/o utils
mindspore-lab/mindaudio
A toolbox of audio models and algorithms based on MindSpore