Marc Delcroix

The paper proposes a method to train speaker diarization models to produce tight predictions using loose labels.

The paper proposes a method called Close-to-Distant microphone Projection (C2D projection) to generate paired data from real recordings for training neural networks for speech enhancement.

This paper proposes a novel non-autoregressive (NAR) decoding framework based on minimum Bayes' risk (MBR) for speech recognition, which outperforms previous NAR decoding and runs faster than autoregressive decoding.

This paper introduces the REAL-TSE Challenge, a satellite challenge on target speaker extraction from real conversational recordings, and describes its task definition, datasets, evaluation protocol, and submitted systems.