Lightweight Adaptive Deep Learning for Efficient Real-Time Speech Enhancement on Edge Devices
IEEE Transactions on Consumer Electronics · 2025-11 · 71 (4) · 12086-12095
DOI: 10.1109/tce.2025.3598007 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Consumer Electronics, 2025-11. Verfasst von Fazal E. Wahab und 4 weiteren Autor:innen. Die Arbeit wurde bisher 16-mal zitiert (Crossref) und nennt 60 Literaturangaben. Gefördert von: National Natural Science Foundation of China, Taif University, Saudi Arabia, for supporting this work, U.K. Engineering and Physical Sciences Research Council.
Abstract (Originalsprache)
Deep learning has significantly advanced speech enhancement (SE) by exploiting hierarchical representations to model complex speech patterns. However, deploying these models on resource-constrained edge devices remains challenging due to computational limitations and real-time processing requirements. Convolutional neural networks (CNNs) face challenges due to frequency translation equivariance, which reduces their sensitivity to frequency-specific features essential for speech-noise separation. Transformer-based SE models are effective at capturing global dependencies but are computationally expensive and less suitable for low-latency edge processing. This study proposes an efficient encoder-decoder architecture optimized for SE on edge devices to address these challenges. The model integrates adaptive frequency-aware gated convolution (AFAGC) in the encoder and a Ginformer-based bottleneck, ensuring robust real-time performance with minimal computational overhead. The encoder incorporates adaptive frequency band positional encoding to mitigate translation equivariance, while gated convolution selectively reweights frequency components to emphasize speech-relevant features. The Ginformer-based bottleneck uses low-rank projections to reduce self-attention complexity and an SRU-based temporal gating to enhance noise adaptation and computational efficiency. Evaluation on the VoiceBank+DEMAND dataset demonstrates that the proposed model outperforms recent SE models, achieving a PESQ of 3.25 and STOI of 95.5%. With only 1.32 million parameters and a real-time factor (RTF) of 0.14, it delivers high-quality speech enhancement suitable for real-time deployment on edge devices.
Förderung
- National Natural Science Foundation of China
- Taif University, Saudi Arabia, for supporting this work
- U.K. Engineering and Physical Sciences Research Council
Zitieren
APA
Wahab, F. E., Ye, Z., Saleem, N., Bourouis, S., & Hussain, A. (2025). Lightweight Adaptive Deep Learning for Efficient Real-Time Speech Enhancement on Edge Devices. IEEE Transactions on Consumer Electronics, 71(4), 12086-12095. https://doi.org/10.1109/tce.2025.3598007
Vancouver
Wahab FE, Ye Z, Saleem N, Bourouis S, Hussain A. Lightweight Adaptive Deep Learning for Efficient Real-Time Speech Enhancement on Edge Devices. IEEE Transactions on Consumer Electronics. 2025;71(4):12086-12095. doi:10.1109/tce.2025.3598007
BibTeX
@article{Wahab2025,
author = {Wahab, Fazal E. and Ye, Zhongfu and Saleem, Nasir and Bourouis, Sami and Hussain, Amir},
title = {Lightweight Adaptive Deep Learning for Efficient Real-Time Speech Enhancement on Edge Devices},
journal = {IEEE Transactions on Consumer Electronics},
year = {2025},
volume = {71},
number = {4},
pages = {12086--12095},
doi = {10.1109/tce.2025.3598007}
}
EUVIMED