PnP-Flow: Plug-and-Play Image Restoration with Flow Matching
About
In this paper, we introduce Plug-and-Play (PnP) Flow Matching, an algorithm for solving imaging inverse problems. PnP methods leverage the strength of pre-trained denoisers, often deep neural networks, by integrating them in optimization schemes. While they achieve state-of-the-art performance on various inverse problems in imaging, PnP approaches face inherent limitations on more generative tasks like inpainting. On the other hand, generative models such as Flow Matching pushed the boundary in image sampling yet lack a clear method for efficient use in image restoration. We propose to combine the PnP framework with Flow Matching (FM) by defining a time-dependent denoiser using a pre-trained FM model. Our algorithm alternates between gradient descent steps on the data-fidelity term, reprojections onto the learned FM path, and denoising. Notably, our method is computationally efficient and memory-friendly, as it avoids backpropagation through ODEs and trace computations. We evaluate its performance on denoising, super-resolution, deblurring, and inpainting tasks, demonstrating superior results compared to existing PnP algorithms and Flow Matching based state-of-the-art methods.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Super-Resolution | CelebA | -- | 24 | |
| Multi-Coil MRI Reconstruction | fastMRI Brain multi-coil 4x acceleration | SSIM95.1 | 13 | |
| Multi-Coil MRI Reconstruction | fastMRI Brain 8x acceleration multi-coil | SSIM0.913 | 13 | |
| Multi-Coil MRI Reconstruction | CMRxRecon Cardiac T1/T2 Mapping multi-coil 8x acceleration | SSIM92.4 | 13 | |
| Multi-Coil MRI Reconstruction | CMRxRecon Cardiac T1 T2 Mapping 4x acceleration multi-coil | SSIM96.3 | 13 | |
| MRI Reconstruction | CMRxRecon 2023 | Time (ms)938 | 12 | |
| Deblurring | CelebA | PSNR34.51 | 11 | |
| Image Inpainting | PIE-Bench (556 samples) | FID66.8 | 11 | |
| Inpainting | FFHQ 768 x 768 5k samples | FID (Half)20.5 | 11 | |
| Image Inpainting | FFHQ DIV2K (val) | Latency (s)3.1 | 11 |