DSConv+LR: A Minimalist Lightweight Network for Image Super-Resolution
Electronics · 15.06.2026 · 15 (12) · 2637
DOI: 10.3390/electronics15122637 ↗Kurz gefasst
Fachartikel in der Zeitschrift Electronics, online erschienen am 15.06.2026. Verfasst von Qiuxia Hu und 4 weiteren Autor:innen. Die Arbeit nennt 15 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Shaanxi NSFC.
Abstract (Originalsprache)
Deep learning has significantly advanced image super-resolution (SR), yet many state-of-the-art models remain too computationally expensive for resource-constrained devices. This paper demonstrates that a highly parameter-efficient design can achieve comparable performance to the very deep super-resolution network (VDSR) with a tiny fraction of parameters. Starting from the classic VDSR architecture (2016), we systematically evaluate three design choices: depthwise separable convolution (DSConv), Hybrid Attention Transformer (HAT), and a local residual connection (LR). HAT provides no performance gain—an honest negative result supported by controlled experiments (increased training, different reduction ratios, and standard convolution baseline). In contrast, LR alone yields a 0.20 dB improvement without introducing any extra parameters. Consequently, we discard HAT and propose DSConv+LR. Our model contains only 49,217 parameters—about 7.4% of VDSR—yet attains a peak signal-to-noise ratio (PSNR) of 35.21 dB on Set5 (×2), which is 99.7% of VDSR’s performance (35.33 dB). On additional benchmarks (Set14, BSD100, and Urban100), DSConv+LR maintains similar relative performance (within 0.12 dB of VDSR). Perceptual loss (AlexNet features, lower better) is 0.2556, slightly better than VDSR (0.2717). We acknowledge that modern lightweight networks such as cascaded residual attention network (CARN) and information multi-distillation network (IMDN) achieve 2–3 dB higher PSNR at the cost of 9–14× more parameters. This work advocates a minimalist approach while honestly reporting both its strengths and limitations.
Förderung
- Shaanxi NSFC
Zitieren
APA
Hu, Q., Tian, J., Jiang, G., Xue, S., & Wang, J. (2026). DSConv+LR: A Minimalist Lightweight Network for Image Super-Resolution. Electronics, 15(12), 2637. https://doi.org/10.3390/electronics15122637
Vancouver
Hu Q, Tian J, Jiang G, Xue S, Wang J. DSConv+LR: A Minimalist Lightweight Network for Image Super-Resolution. Electronics. 2026;15(12):2637. doi:10.3390/electronics15122637
BibTeX
@article{Hu2026,
author = {Hu, Qiuxia and Tian, Jie and Jiang, Guangyi and Xue, Shan and Wang, Jingxuan},
title = {DSConv+LR: A Minimalist Lightweight Network for Image Super-Resolution},
journal = {Electronics},
year = {2026},
volume = {15},
number = {12},
pages = {2637},
doi = {10.3390/electronics15122637}
}
EUVIMED