Asteroid: the PyTorch-based audio source separation toolkit for researchers - Archive ouverte HAL Access content directly
Conference Papers Year :

Asteroid: the PyTorch-based audio source separation toolkit for researchers

(1) , (2) , (1) , (1) , (3) , (4) , (1) , (5) , (1) , (6) , (7) , (8) , (1) , (1)
1
2
3
4
5
6
7
8
Fabian-Robert Stöter

Abstract

This paper describes Asteroid, the PyTorch-based audio source separation toolkit for researchers. Inspired by the most successful neural source separation systems, it provides all neu-ral building blocks required to build such a system. To improve reproducibility, Kaldi-style recipes on common audio source separation datasets are also provided. This paper describes the software architecture of Asteroid and its most important features. By showing experimental results obtained with Asteroid's recipes, we show that our implementations are at least on par with most results reported in reference papers. The toolkit is publicly available at github.com/mpariente/asteroid.
Fichier principal
Vignette du fichier
old_main.pdf (942.46 Ko) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-02962964 , version 1 (09-10-2020)

Identifiers

  • HAL Id : hal-02962964 , version 1

Cite

Manuel Pariente, Samuele Cornell, Joris Cosentino, Sunit Sivasankaran, Efthymios Tzinis, et al.. Asteroid: the PyTorch-based audio source separation toolkit for researchers. Interspeech 2020, Oct 2020, Shanghai, China. ⟨hal-02962964⟩
430 View
462 Download

Share

Gmail Facebook Twitter LinkedIn More