DEALing with Image Reconstruction: Deep Attentive Least Squares
M. Pourya, E. Kobler, M. Unser, S. Neumayer
Proceedings of the Forty-Second International Conference on Machine Learning (ICML'25), Vancouver BC, Canada, July 13-19, 2025, 20 p.
State-of-the-art image reconstruction often relies on complex, abundantly parameterized deep architectures. We propose an alternative: a data-driven reconstruction method inspired by the classic Tikhonov regularization. Our approach iteratively refines intermediate reconstructions by solving a sequence of quadratic problems. These updates have two key components: (i) learned filters to extract salient image features; and (ii) an attention mechanism that locally adjusts the penalty of the filter responses. Our method matches leading plug-and-play and learned regularizer approaches in performance while offering interpretability, robustness, and convergent behavior. In effect, we bridge traditional regularization and deep learning with a principled reconstruction approach.
@INPROCEEDINGS(http://bigwww.epfl.ch/publications/pourya2501.html,
AUTHOR="Pourya, M. and Kobler, E. and Unser, M. and Neumayer, S.",
TITLE="{DEALing} with Image Reconstruction: {D}eep Attentive Least
Squares",
BOOKTITLE="Proceedings of the Forty-Second International Conference on
Machine Learning ({ICML'25})",
YEAR="2025",
editor="",
volume="267",
series="",
pages="",
address="Vancouver BC, Canada",
month="July 13-19,",
organization="",
publisher="",
note="")