mlsanityData to 7 Oct 2026

Unsupervised Multi-channel Speech Dereverberation via Diffusion

arXiv 2508.02071 ↗First posted 4 Aug 2025cs.SD eess.AScited by 0

On the map: Diffusion models → Audio, speech and music

Abstract (arXiv)

We consider the problem of multi-channel single-speaker blind dereverberation, where multi-channel mixtures are used to recover the clean anechoic speech. To solve this problem, we propose USD-DPS, {U}nsupervised {S}peech {D}ereverberation via {D}iffusion {P}osterior {S}ampling. USD-DPS uses an unconditional clean speech diffusion model as a strong prior to solve the problem by posterior sampling. At each diffusion sampling step, we estimate all microphone channels' room impulse responses (RIRs), which are further used to enforce a multi-channel mixture consistency constraint for diffusion guidance. For multi-channel RIR estimation, we estimate reference-channel RIR by optimizing RIR parameters of a sub-band RIR signal model, with the Adam optimizer. We estimate non-reference channels' RIRs analytically using forward convolutive prediction (FCP). We found that this combination provides a good balance between sampling efficiency and RIR prior modeling, which shows superior performance among unsupervised dereverberation approaches. An audio demo page is provided in https://usddps.github.io/USDDPS_demo/.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Audio, speech and music, one of 14 directions in Diffusion models. Builds on 26 of its 39 references that are in the map. First posted on arXiv 4 Aug 2025.

Builds on

26 of 39 references.

Closest ancestors · 9

Same direction, newest first

  1. 2025ArrayDPS: Unsupervised Blind Speech Separation with a Diffusion Prior8 May
  2. 2025Unsupervised Estimation of Nonlinear Audio Effects: Comparing Diffusion-Based and Adversarial approaches7 Apr
  3. 2024HRTF Estimation using a Score-based Prior2 Oct
  4. 2024Unsupervised Blind Joint Dereverberation and Room Acoustics Estimation with Diffusion Models14 Aug
  5. 2023Zero-Shot Duet Singing Voices Separation with Diffusion Models13 Nov
  6. 2023Diffusion Posterior Sampling for Informed Single-Channel Dereverberation21 Jun
  7. 2023Mousai: Text-to-Music Generation with Long-Context Latent Diffusion27 Jan
  8. 2022Unsupervised vocal dereverberation with diffusion-based generative models8 Nov
  9. 2022Solving Audio Inverse Problems with a Diffusion Model27 Oct

Other references · 17

  1. 2020Denoising Diffusion Probabilistic Modelscited by 20,851
  2. 2020Score-Based Generative Modeling through Stochastic Differential Equationscited by 8,964
  3. 2022Elucidating the Design Space of Diffusion-Based Generative Modelscited by 2,994
  4. 2022Diffusion Posterior Sampling for General Noisy Inverse Problemscited by 1,375
  5. 2019LibriTTS: A Corpus Derived from LibriSpeech for Text-to-Speechcited by 639
  6. 2018SDR - half-baked or well done?cited by 420
  7. 2022Parallel Diffusion Models of Operator and Image for Blind Inverse Problemscited by 113
  8. 2022TF-GridNet: Integrating Full- and Sub-Band Modeling for Speech Separationcited by 62
  9. 2023Fast Diffusion EM: a diffusion model for blind inverse problems with application to deconvolutioncited by 36
  10. 2023Blind Audio Bandwidth Extension: A Diffusion-Based Zero-Shot Approachcited by 21
  11. 2020Multi-Microphone Complex Spectral Mapping for Speech Dereverberationcited by 8
  12. 2021Convolutive Prediction for Monaural Speech Dereverberation and Noisy-Reverberant Speaker Separationcited by 7
5 more
  1. 2024USDnet: Unsupervised Speech Dereverberation via Neural Forward Filteringcited by 6
  2. 2021Multi-channel Narrow-band Deep Speech Separation with Full-band Permutation Invariant Trainingcited by 4
  3. 2023UNSSOR: Unsupervised Neural Speech Separation by Leveraging Over-determined Training Mixturescited by 4
  4. 2021Convolutive Prediction for Reverberant Speech Separationcited by 3
  5. 2022Multichannel Speech Separation with Narrow-band Conformercited by 3

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2024BUDDy: Single-Channel Blind Unsupervised Dereverberation with Diffusion Models
  2. 2025On the Application of Diffusion Models for Simultaneous Denoising and Dereverberation
  3. 2026Unified Diffusion Refinement for Multi-Channel Speech Enhancement and Separation
  4. 2022Speech Enhancement and Dereverberation with Diffusion-based Generative Models
  5. 2025Unsupervised Single-Channel Audio Separation with Diffusion Source Priors
  6. 2026VM-ARRAYDPS: Virtual Microphone Augmented Diffusion Posterior Sampling for Unsupervised Blind Speech Separation
  7. 2026Diffusion-based Frameworks for Unsupervised Speech Enhancement
  8. 2026SSNAPS: Audio-Visual Separation of Speech and Background Noise with Diffusion Inverse Sampling