asteroid-team/asteroid
A PyTorch toolkit for researchers to experiment with deep learning-based audio source separation models and architectures.
Not currently ranked — collecting fresh signals.
star history
Asteroid is a research-focused toolkit that provides PyTorch implementations of audio source separation models, supporting various datasets and architectures. It includes pre-trained models and recipes to reproduce published papers. The toolkit enables researchers to quickly experiment with common audio separation benchmarks and develop new separation models.
Frequently asked
- What is asteroid-team/asteroid?
- A PyTorch toolkit for researchers to experiment with deep learning-based audio source separation models and architectures.
- Is asteroid open source?
- Yes — asteroid-team/asteroid is open source, released under the MIT license.
- What language is asteroid written in?
- asteroid-team/asteroid is primarily written in Python.
- How popular is asteroid?
- asteroid-team/asteroid has 2.6k stars on GitHub.
- Where can I find asteroid?
- asteroid-team/asteroid is on GitHub at https://github.com/asteroid-team/asteroid.