Content-Aware Transformer Network for Image Compressed Sensing
IEEE Internet of Things Journal · 01.06.2026 · 13 (11) · 25274-25285
DOI: 10.1109/jiot.2026.3676729 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Internet of Things Journal, erschienen am 01.06.2026. Verfasst von Dawei Song und 4 weiteren Autor:innen. Die Arbeit nennt 55 Literaturangaben. Gefördert von: Major Discipline Academic and Technical Leaders Training Program of Jiangxi Province, Postdoctoral Fellowship Program of CPSF, China Postdoctoral Science Foundation.
Abstract (Originalsprache)
Accurate image reconstruction from low-sampling measurements remains a crucial challenge in image compressed sensing (CS). Image CS has attracted increasing research interest, with deep learning approaches being extensively studied. Deep learning-based methods reduce computational complexity at the inference stage compared with traditional iterative solvers and can improve reconstruction quality. In particular, methods built on transformer architectures have achieved remarkable success in tasks involving image reconstruction and restoration. However, as the depth of the model increases, some content information may be attenuated during feature propagation. To alleviate the issue, this article proposes a content-aware transformer network (CATNet) for image CS. It focuses on content feed-forward and channel self-attention to effectively obtain sufficient feature representation. In the initial reconstruction stage, this article employs trainable sensing matrix to adaptively capture sparse structures from the image, which helps improve measurement efficiency. In the deep reconstruction stage, this article introduces the content-focused feed-forward network (CFFN) to strengthen the effective content features. Besides, this article employs the channel-focused self-attention (CFSA) to enhance global features and integrate convolution to refine local details. Finally, this article adopts a multistage supervision to ensure sufficient training. Comprehensive experiments conducted on five widely used datasets demonstrate the strong performance and superiority of CATNet over nine representative methods, especially under extremely low sampling rates. At CS ratio of 0.05, CATNet achieves competitive results, with peak signal-to-noise ratio (PSNR)/structural similarity index measure (SSIM) values of 22.25/0.6907 on Set5, 21.91/0.6623 on Set11, 21.99/0.6664 on Set14, 23.97/0.6964 on BSD68, and 20.61/0.5585 on Urban100. Moreover, the method has been successfully implemented in the self-developed single-pixel imaging (SPI) optical system, further demonstrating its applicability.
Förderung
- Major Discipline Academic and Technical Leaders Training Program of Jiangxi Province
- Postdoctoral Fellowship Program of CPSF
- China Postdoctoral Science Foundation
- National Natural Science Foundation of China
Zitieren
APA
Song, D., Yan, Q., Yang, J., Han, P., & Li, X. (2026). Content-Aware Transformer Network for Image Compressed Sensing. IEEE Internet of Things Journal, 13(11), 25274-25285. https://doi.org/10.1109/jiot.2026.3676729
Vancouver
Song D, Yan Q, Yang J, Han P, Li X. Content-Aware Transformer Network for Image Compressed Sensing. IEEE Internet of Things Journal. 2026;13(11):25274-25285. doi:10.1109/jiot.2026.3676729
BibTeX
@article{Song2026,
author = {Song, Dawei and Yan, Qiurong and Yang, Jian and Han, Pengfei and Li, Xuelong},
title = {Content-Aware Transformer Network for Image Compressed Sensing},
journal = {IEEE Internet of Things Journal},
year = {2026},
volume = {13},
number = {11},
pages = {25274--25285},
doi = {10.1109/jiot.2026.3676729}
}
EUVIMED