Metraipolis › archive › dc_a2b761db45c5

SlimDiffuSE: Towards Efficient Diffusion-Based Speech Enhancement using Slimmable Networks — arXiv:2608.21188

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.21188
License (as archived)
CC0-1.0
Archived
2026-10-18T11:35:37.516Z
SHA-256 of the archived text
f87f17a69e85376001b7cad548d12cde3c030c9fe5e719e65138b077be1bdba9
Stable URL of this page
https://metraipolis.com/documento/dc_a2b761db45c5

How to cite

SlimDiffuSE: Towards Efficient Diffusion-Based Speech Enhancement using Slimmable Networks — arXiv:2608.21188 — arXiv (en), https://arxiv.org/abs/2608.21188 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_a2b761db45c5 (Metraipolis archive dc_a2b761db45c5, sha256 f87f17a69e85376001b7cad548d12cde3c030c9fe5e719e65138b077be1bdba9)
BibTeX
@misc{dca2b761db45c5,
  title = {SlimDiffuSE: Towards Efficient Diffusion-Based Speech Enhancement using Slimmable Networks — arXiv:2608.21188},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.21188},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_a2b761db45c5; sha256 f87f17a69e85376001b7cad548d12cde3c030c9fe5e719e65138b077be1bdba9},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.21188. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Diffusion-based models are emerging in the speech enhancement domain and are achieving state-of-the-art performance across various benchmark datasets. A major downside of diffusion models is that data generation requires many evaluations of a typically large neural network, which results in high overall complexity. In this work, we propose a slimmable diffusion model that employs adaptive network widths throughout the data generation process to reduce computational cost. By using a greedy search algorithm to optimize the network width schedule, our method achieves performance comparable to baseline diffusion models with significantly reduced computational complexity. Notably, our approach reduces the computational complexity by up to $87.5\%$ without a significant drop in objective metrics, such as perceptual evaluation of speech quality (PESQ) and SI-SDR.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-24 in eess.AS).
Categorie arXiv: Audio and Speech Processing (eess.AS).
Pagina del lavoro: https://arxiv.org/abs/2608.21188 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-18T11:29:11.004Z dalla pagina https://arxiv.org/catchup/eess.AS/2026-08-24?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).