CoFi-Lite: Pushing the Limits of Ultra-Lightweight Speech Enhancement
IEEE Signal Processing Letters · 2026 · 33 · 2954-2958
DOI: 10.1109/lsp.2026.3712291 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Signal Processing Letters, 2026. Verfasst von Leyan Yang und 4 weiteren Autor:innen. Die Arbeit nennt 34 Literaturangaben. Gefördert von: National Natural Science Foundation of China.
Abstract (Originalsprache)
Ultra-lightweight models are essential for the deployment of deep learning-based speech enhancement algorithms on edge devices. Although recent approaches have achieved a certain balance between computational complexity and performance, pushing the complexity limits further demands more sophisticated designs. In this letter, we propose CoFi-Lite, a highly efficient model that decouples spectral modeling into coarse- and fine-grained streams. By leveraging two parallel and symmetric encoder-decoder paths, it simultaneously extracts full-band envelopes and low-frequency details for complementary enhancement. In addition, a novel Cross-Path Fusion (CPF) module is introduced to bridge the distinct paths, facilitating efficient feature interaction. Remarkably, CoFi-Lite requires extremely low computational resources, featuring only 12.87 M MACs/s and 83.12 k parameters. Experimental results demonstrate that our proposed model outperforms the ultra-lightweight baseline GTCRN while requiring only 40.26% of its computational complexity. Its scaled-up variant also delivers performance on par with that of the SOTA ultra-lightweight model AdaptCRN alongside a 19.34% reduction in computational cost.
Förderung
- National Natural Science Foundation of China
Zitieren
APA
Yang, L., Wang, D., Rong, X., Zhao, J., & Lu, J. (2026). CoFi-Lite: Pushing the Limits of Ultra-Lightweight Speech Enhancement. IEEE Signal Processing Letters, 33, 2954-2958. https://doi.org/10.1109/lsp.2026.3712291
Vancouver
Yang L, Wang D, Rong X, Zhao J, Lu J. CoFi-Lite: Pushing the Limits of Ultra-Lightweight Speech Enhancement. IEEE Signal Processing Letters. 2026;33:2954-2958. doi:10.1109/lsp.2026.3712291
BibTeX
@article{Yang2026,
author = {Yang, Leyan and Wang, Dahan and Rong, Xiaobin and Zhao, Jiadong and Lu, Jing},
title = {CoFi-Lite: Pushing the Limits of Ultra-Lightweight Speech Enhancement},
journal = {IEEE Signal Processing Letters},
year = {2026},
volume = {33},
pages = {2954--2958},
doi = {10.1109/lsp.2026.3712291}
}
EUVIMED