Alexandre Défossez

Research Scientist

I have been a Research Scientist at FAIR Paris since October 2020.

Before that, I completed my PhD between FAIR and INRIA in Paris under the supervision of Léon Bottou, Nicolas Usunier, and Francis Bach. I work mostly on deep learning based signal processing, with applications to source separation or the analysis of brain activity data. I also enjoy working on theoretical aspects of machine learning, and in particular non convex optimization.


Deep learning based signal processing, source separation, audio synthesis, neuroscience, non convex optimization

Latest Publications

Interspeech - October 25, 2020

Real Time Speech Enhancement in the Waveform Domain

Alexandre Défossez, Gabriel Synnaeve, Yossi Adi

NIPS 2018 - October 26, 2018

SING: Symbol-to-Instrument Neural Generator

Alexandre Défossez, Neil Zeghidour, Nicolas Usunier, Leon Bottou, Francis Bach