Skip to main content
archive
Search Submit Donate Log in
Press Enter to search · Advanced search

Electrical Engineering and Systems Science > Audio and Speech Processing

arXiv:2305.05857 (eess)
[Submitted on 10 May 2023 (v1), last revised 10 Feb 2026 (this version, v3)]

Title:Diffusion-based Signal Refiner for Speech Enhancement and Separation

Authors:Masato Hirano, Ryosuke Sawata, Naoki Murata, Shusuke Takahashi, Yuki Mitsufuji
View a PDF of the paper titled Diffusion-based Signal Refiner for Speech Enhancement and Separation, by Masato Hirano and 4 other authors
View PDF HTML (experimental)
Abstract:Although recent speech processing technologies have achieved significant improvements in objective metrics, there still remains a gap in human perceptual quality. This paper proposes Diffiner, a novel solution that utilizes the powerful generative capability of diffusion models' prior distributions to address this fundamental issue. Diffiner leverages the probabilistic generative framework of diffusion models and learns natural prior distributions of clean speech to convert outputs from existing speech processing systems into perceptually natural high-quality audio. In contrast to conventional deterministic approaches, our method simultaneously analyzes both the original degraded speech and the pre-processed speech to accurately identify unnatural artifacts introduced during processing. Then, through the iterative sampling process of the diffusion model, these degraded portions are replaced with perceptually natural and high-quality speech segments. Experimental results indicate that Diffiner can recover a clearer harmonic structure of speech, which is shown to result in improved perceptual quality w.r.t. several metrics as well as in a human listening test. This highlights Diffiner's efficacy as a versatile post-processor for enhancing existing speech processing pipelines.
Comments: Accepted to IEEE/ACM TASLP. The first two authors contributed equally. Code: this https URL
Subjects: Audio and Speech Processing (eess.AS); Sound (cs.SD)
Cite as: arXiv:2305.05857 [eess.AS]
  (or arXiv:2305.05857v3 [eess.AS] for this version)
  https://doi.org/10.48550/arXiv.2305.05857
arXiv-issued DOI via DataCite
Related DOI: https://doi.org/10.1109/TASLPRO.2026.3663910
DOI(s) linking to related resources

Submission history

From: Masato Hirano [view email]
[v1] Wed, 10 May 2023 03:06:16 UTC (14,352 KB)
[v2] Fri, 12 May 2023 05:02:05 UTC (14,352 KB)
[v3] Tue, 10 Feb 2026 03:58:36 UTC (716 KB)
Full-text links:

Access Paper:

    View a PDF of the paper titled Diffusion-based Signal Refiner for Speech Enhancement and Separation, by Masato Hirano and 4 other authors
  • View PDF
  • HTML (experimental)
  • TeX Source
view license

Current browse context:

eess.AS
< prev   |   next >
new | recent | 2023-05
Change to browse by:
cs
cs.SD
eess

References & Citations

  • NASA ADS
  • Google Scholar
  • Semantic Scholar
Loading...

BibTeX formatted citation

Data provided by:

Bookmark

BibSonomy Reddit

Bibliographic and Citation Tools

Bibliographic Explorer (What is the Explorer?)
Connected Papers (What is Connected Papers?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)

Code, Data and Media Associated with this Article

alphaXiv (What is alphaXiv?)
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Hugging Face (What is Huggingface?)
ScienceCast (What is ScienceCast?)

Demos

Replicate (What is Replicate?)
Hugging Face Spaces (What is Spaces?)
TXYZ.AI (What is TXYZ.AI?)

Recommenders and Search Tools

Influence Flower (What are Influence Flowers?)
CORE Recommender (What is CORE?)
  • Author
  • Venue
  • Institution
  • Topic

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)
We gratefully acknowledge support from our major funders, member institutions, , and all contributors.
About · Help · Contact · Subscribe · Copyright · Privacy · Accessibility · Operational Status (opens in new tab)
Major funding support from
Simons Foundation Simons Foundation International Schmidt Sciences