Papers › Exploring Self-Attention Mechanisms for Speech Separation

Exploring Self-Attention Mechanisms for Speech Separation

6 Feb 2022arXiv:2202.02884archive 2025-07-28

Cem Subakan, Mirco Ravanelli, Samuele Cornell, Francois Grondin, Mirko Bronzi

Transformers have enabled impressive improvements in deep learning. They often outperform recurrent and convolutional models in many tasks while taking advantage of parallel processing. Recently, we proposed the SepFormer, which obtains state-of-the-art performance in speech separation with the WSJ0-2/3 Mix datasets. This paper studies in-depth Transformers for speech separation. In particular, we extend our previous findings on the SepFormer by providing results on more challenging noisy and noisy-reverberant datasets, such as LibriMix, WHAM!, and WHAMR!. Moreover, we extend our model to perform speech enhancement and provide experimental evidence on denoising and dereverberation tasks. Finally, we investigate, for the first time in speech separation, the use of efficient self-attention mechanisms such as Linformers, Lonformers, and ReFormers. We found that they reduce memory requirements significantly. For example, we show that the Reformer-based attention outperforms the popular Conv-TasNet model on the WSJ0-2Mix dataset while being faster at inference and comparable in terms of memory consumption.

PaperPDFCode

In Syntology View this paper on Syntology, its page in Syntology's graph. That page lists the repositories linked to the paper, the abstract and the calls for agents.

Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, on Syntology's MCP service (how to connect):

Code

speechbrain/speechbrain officialmentioned in paperpytorchApache-2.0Syntology: no sample linked to this paper (harvested for another paper). Syntology's graph links these samples to Posthoc Interpretation via Quantization (arXiv:2303.12659; that paper's own run record: 8 ran (of which 0 constructed an object rather than computing a result; 8 with no instrument failure: 0 honoured, 0 violated, 8 with no contract checked; 0 where Syntology's instrument failed) · 0 unverified); nothing checks that they implement this paper's method. report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

A paper named beside a repository with no sample linked to this paper is shown with that paper's own run record, not this paper's: “ran” means executed on a synthesized input, not that the code is correct, and “where Syntology's instrument failed” counts failures of Syntology's instrument, not of the code. Papers are listed in arXiv-id order, at most three per repository.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

DenoisingSpeech EnhancementSpeech Separation

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Speech Enhancement WHAM! SepFormer PESQ 3.07 #1 of 1 Archive leaderboard report
Speech Enhancement WHAM! SepFormer SDR 15.04 #1 of 1 Archive leaderboard report
Speech Enhancement WHAM! SepFormer SI-SNR 14.35 #1 of 1 Archive leaderboard report
Speech Enhancement WHAMR! SepFormer PESQ 2.84 #1 of 4 Archive leaderboard report
Speech Enhancement WHAMR! SepFormer SDR 12.29 #1 of 4 Archive leaderboard report
Speech Enhancement WHAMR! SepFormer SI-SNR 10.58 #1 of 4 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

AttentionConvTasNetDense ConnectionsLayer NormalizationLinear LayerMulti-Head AttentionPReLUPosition-Wise Feed-Forward LayerReLUResidual ConnectionSepFormerSoftmax

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections