Papers › Separate And Diffuse: Using a Pretrained Diffusion Model for Improving Source Separation

Separate And Diffuse: Using a Pretrained Diffusion Model for Improving Source Separation

25 Jan 2023arXiv:2301.10752archive 2025-07-28

Shahar Lutati, Eliya Nachmani, Lior Wolf

The problem of speech separation, also known as the cocktail party problem, refers to the task of isolating a single speech signal from a mixture of speech signals. Previous work on source separation derived an upper bound for the source separation task in the domain of human speech. This bound is derived for deterministic models. Recent advancements in generative models challenge this bound. We show how the upper bound can be generalized to the case of random generative models. Applying a diffusion model Vocoder that was pretrained to model single-speaker voices on the output of a deterministic separation model leads to state-of-the-art separation results. It is shown that this requires one to combine the output of the separation model with that of the diffusion model. In our method, a linear combination is performed, in the frequency domain, using weights that are inferred by a learned model. We show state-of-the-art results on 2, 3, 5, 10, and 20 speakers on multiple benchmarks. In particular, for two speakers, our method is able to surpass what was previously considered the upper performance bound.

PaperPDF

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

Code

No code repository is listed for this paper in the archive or in Syntology's graph.

Code Syntology ran Syntology

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

Tasks

Audio Source SeparationGeneralization BoundsMulti-Speaker Source SeparationSpeech Separation

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Speech Separation Libri10Mix Separate And Diffuse SI-SDRi 9 #1 of 3 Archive leaderboard report
Speech Separation Libri20Mix Separate And Diffuse SI-SDRi 5.2 #1 of 2 Archive leaderboard report
Speech Separation Libri2Mix Separate And Diffuse SI-SDRi 21.5 #4 of 10 Archive leaderboard report
Speech Separation Libri5Mix Separate And Diffuse SI-SDRi 14.2 #1 of 4 Archive leaderboard report
Speech Separation WSJ0-2mix Separate And Diffuse SI-SDRi 23.9 #7 of 40 Archive leaderboard report
Speech Separation WSJ0-3mix Separate And Diffuse SI-SDRi 20.9 #4 of 9 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

Introduced by this paper: Separate And Diffuse

AttentionDense ConnectionsDiffusionLayer NormalizationLinear LayerMulti-Head AttentionPReLUPosition-Wise Feed-Forward LayerReLUResidual ConnectionSepFormerSeparate And DiffuseSoftmax

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