Alexandre Defossez

Paris, France

Alexandre Defossez joined FAIR Paris as a research scientist in October 2020. He works mostly on deep learning based digital signal processing, with applications to audio source separation, audio compression, and brain signal decoding. He also enjoys working on more theoretical aspect of deep learning. He completed his PhD between FAIR and INRIA in Paris under the supervision of Léon Bottou, Nicolas Usunier, and Francis Bach.

Alexandre's Work

Alexandre's Publications

October 25, 2020

RESEARCH

Real Time Speech Enhancement in the Waveform Domain

Alexandre Defossez, Gabriel Synnaeve, Yossef Mordechay Adi

October 25, 2020

October 26, 2018

RESEARCH

SING: Symbol-to-Instrument Neural Generator

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

October 26, 2018