DEALing with Image Reconstruction: Deep Attentive Least Squares
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2025
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866916600827871232 |
|---|---|
| author | Pourya, Mehrsa Kobler, Erich Unser, Michael Neumayer, Sebastian |
| author_facet | Pourya, Mehrsa Kobler, Erich Unser, Michael Neumayer, Sebastian |
| contents | State-of-the-art image reconstruction often relies on complex, highly 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 filter responses. Our method achieves performance on par with leading plug-and-play and learned regularizer approaches while offering interpretability, robustness, and convergent behavior. In effect, we bridge traditional regularization and deep learning with a principled reconstruction approach. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_04079 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | DEALing with Image Reconstruction: Deep Attentive Least Squares Pourya, Mehrsa Kobler, Erich Unser, Michael Neumayer, Sebastian Image and Video Processing Computer Vision and Pattern Recognition Machine Learning State-of-the-art image reconstruction often relies on complex, highly 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 filter responses. Our method achieves performance on par with leading plug-and-play and learned regularizer approaches while offering interpretability, robustness, and convergent behavior. In effect, we bridge traditional regularization and deep learning with a principled reconstruction approach. |
| title | DEALing with Image Reconstruction: Deep Attentive Least Squares |
| topic | Image and Video Processing Computer Vision and Pattern Recognition Machine Learning |
| url | https://arxiv.org/abs/2502.04079 |