@inproceedings{f7642f3523bf47ac8e1609f4bf6f5a1e,
title = "Multiple Choice Learning for Efficient Speech Separation with Many Speakers",
abstract = "Training speech separation models in the supervised setting raises a permutation problem: finding the best assignation between the model predictions and the ground truth separated signals. This inherently ambiguous task is customarily solved using Permutation Invariant Training (PIT). In this article, we instead consider using the Multiple Choice Learning (MCL) framework, which was originally introduced to tackle ambiguous tasks. We demonstrate experimentally on the popular WSJ0-mix and LibriMix benchmarks that MCL matches the performances of PIT, while being computationally advantageous. This opens the door to a promising research direction, as MCL can be naturally extended to handle a variable number of speakers, or to tackle speech separation in the unsupervised setting.",
keywords = "Librimix, PIT, WSJ0-mix, cocktail party, multiple choice learning, speech separation",
author = "David Perera and Fran{\c c}ois Derrida and Th{\'e}o Mariotte and Ga{\"e}l Richard and Slim Essid",
note = "Publisher Copyright: {\textcopyright} 2025 IEEE.; 2025 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2025 ; Conference date: 06-04-2025 Through 11-04-2025",
year = "2025",
month = jan,
day = "1",
doi = "10.1109/ICASSP49660.2025.10888528",
language = "English",
series = "ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
editor = "Rao, \{Bhaskar D\} and Isabel Trancoso and Gaurav Sharma and Mehta, \{Neelesh B.\}",
booktitle = "2025 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2025 - Proceedings",
}