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.

Stats

• citations1
• S2 citations—
• github stars0
• HF likes0
• heat score2.26
• arxiv keyperera2024multiple

Related papers

• Permutation Invariant Training Of Deep Models For Speaker-independent Multi-talker Speech Separation (2016)0.00
• Interrupted And Cascaded Permutation Invariant Training For Speech Separation (2019)4.52
• Single-channel Speech Separation Using Soft-minimum Permutation Invariant Training (2021)2.26
• Probabilistic Permutation Invariant Training For Speech Separation (2019)7.81
• Single-channel Multi-talker Speech Recognition With Permutation Invariant Training (2017)12.10
• Separating Long-form Speech With Group-wise Permutation Invariant Training (2021)4.52
• Recognizing Multi-talker Speech With Permutation Invariant Training (2017)12.81
• Recursive Speech Separation For Unknown Number Of Speakers (2019)12.93