| Paper ID | SS-2.4 |
| Paper Title |
A PLUG-AND-PLAY DEEP IMAGE PRIOR |
| Authors |
Zhaodong Sun, Fabian Latorre, Thomas Sanchez, Volkan Cevher, Ecole Polytechnique Fédérale de Lausanne (EPFL), Switzerland |
| Session | SS-2: Deep Learning Methods for Solving Linear Inverse Problems |
| Location | Gather.Town |
| Session Time: | Tuesday, 08 June, 14:00 - 14:45 |
| Presentation Time: | Tuesday, 08 June, 14:00 - 14:45 |
| Presentation |
Poster
|
| Topic |
Special Sessions: Deep Learning Methods for Solving Linear Inverse Problems |
| IEEE Xplore Open Preview |
Click here to view in IEEE Xplore |
| Virtual Presentation |
Click here to watch in the Virtual Conference |
| Abstract |
Deep image priors (DIP) offer a novel approach for the regularization that leverages the inductive bias of a deep convolutional architecture in inverse problems. However, the quality of DIP approaches often degrades when the number of iterations exceeds a certain threshold due to overfitting. To mitigate this effect, this work incorporates a plug-and-play prior scheme which can accommodate additional regularization steps within a DIP framework. Our modification is achieved using an augmented Lagrangian formulation of the problem, and is solved using an Alternating Direction Method of Multipliers (ADMM) variant, which can capture existing DIP approaches as a special case. We show experimentally that our ADMM-based DIP pairing outperforms competitive baselines in PSNR while exhibiting less overfitting. |
