Application of Deep Machine Learning in Compressed Sensing Reconstruction of Shift-Invariant Spaces
Electronics · 07.05.2026 · 15 (10) · 1975
DOI: 10.3390/electronics15101975 ↗Kurz gefasst
Fachartikel in der Zeitschrift Electronics, online erschienen am 07.05.2026. Verfasst von Chenyu Ling und 3 weiteren Autor:innen. Die Arbeit nennt 34 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Graduate Education Reform Project Fund of Chengdu University.
Abstract (Originalsprache)
This paper proposes a structure-constrained deep reconstruction framework for compressed sensing in shift-invariant spaces (SISs). The reconstruction is formulated as an inverse operator estimation problem derived from the matrix factorization H(ω)=W(ω)A and approximated using a hybrid CNN–Transformer architecture. Residual dilated convolutions capture localized signal structures, while the Transformer module models global frequency-domain dependencies. A variational inference-inspired regularization mechanism is incorporated to implicitly learn sparsity-aware priors. Experiments on both synthetic SIS signals and real-world ECG data demonstrate consistent improvements over classical optimization-based algorithms (ISTA, OMP) and a deep unfolding baseline (ISTA-Net+). At a 30% sampling rate, the proposed method achieves a PSNR of 35.46 dB. The feed-forward design eliminates iterative reconstruction, achieving a GPU inference time of 0.85 ms per signal.
Förderung
- Graduate Education Reform Project Fund of Chengdu University
Zitieren
APA
Ling, C., Luo, J., Shi, K., & Liu, L. (2026). Application of Deep Machine Learning in Compressed Sensing Reconstruction of Shift-Invariant Spaces. Electronics, 15(10), 1975. https://doi.org/10.3390/electronics15101975
Vancouver
Ling C, Luo J, Shi K, Liu L. Application of Deep Machine Learning in Compressed Sensing Reconstruction of Shift-Invariant Spaces. Electronics. 2026;15(10):1975. doi:10.3390/electronics15101975
BibTeX
@article{Ling2026,
author = {Ling, Chenyu and Luo, Junyi and Shi, Kaibo and Liu, Lusheng},
title = {Application of Deep Machine Learning in Compressed Sensing Reconstruction of Shift-Invariant Spaces},
journal = {Electronics},
year = {2026},
volume = {15},
number = {10},
pages = {1975},
doi = {10.3390/electronics15101975}
}
EUVIMED