Bidirectional Multi-Scale Implicit Neural Representations for Image Deraining (2404.01547v1)
Abstract: How to effectively explore multi-scale representations of rain streaks is important for image deraining. In contrast to existing Transformer-based methods that depend mostly on single-scale rain appearance, we develop an end-to-end multi-scale Transformer that leverages the potentially useful features in various scales to facilitate high-quality image reconstruction. To better explore the common degradation representations from spatially-varying rain streaks, we incorporate intra-scale implicit neural representations based on pixel coordinates with the degraded inputs in a closed-loop design, enabling the learned features to facilitate rain removal and improve the robustness of the model in complex scenarios. To ensure richer collaborative representation from different scales, we embed a simple yet effective inter-scale bidirectional feedback operation into our multi-scale Transformer by performing coarse-to-fine and fine-to-coarse information communication. Extensive experiments demonstrate that our approach, named as NeRD-Rain, performs favorably against the state-of-the-art ones on both synthetic and real-world benchmark datasets. The source code and trained models are available at https://github.com/cschenxiang/NeRD-Rain.
- Not just streaks: Towards ground truth for single image deraining. In ECCV, pages 723–740, 2022.
- Mip-nerf: A multiscale representation for anti-aliasing neural radiance fields. In ICCV, pages 5855–5864, 2021.
- Crossvit: Cross-attention multi-scale vision transformer for image classification. In ICCV, pages 357–366, 2021a.
- Nerv: Neural representations for videos. NeurIPS, 34:21557–21568, 2021b.
- Pre-trained image processing transformer. In CVPR, pages 12299–12310, 2021c.
- Cascaded local implicit transformer for arbitrary-scale super-resolution. In CVPR, pages 18257–18267, 2023a.
- Unpaired deep image deraining using dual contrastive learning. In CVPR, pages 2017–2026, 2022.
- Learning a sparse transformer network for effective image deraining. In CVPR, pages 5896–5905, 2023b.
- Towards unified deep image deraining: A survey and a new benchmark. arXiv preprint arXiv:2310.03535, 2023c.
- Learning continuous image representation with local implicit image function. In CVPR, pages 8628–8638, 2021d.
- Learning implicit fields for generative shape modeling. In CVPR, pages 5939–5948, 2019.
- Rethinking coarse-to-fine approach in single image deblurring. In ICCV, pages 4641–4650, 2021.
- An image is worth 16x16 words: Transformers for image recognition at scale. ICLR, 2020.
- Multiscale vision transformers. In ICCV, pages 6824–6835, 2021.
- Removing rain from single images via a deep detail network. In CVPR, pages 3855–3863, 2017.
- Lightweight pyramid networks for image deraining. IEEE TNNLS, 31(6):1794–1807, 2019.
- Rain streak removal via dual graph convolutional network. In AAAI, pages 1352–1360, 2021.
- Shacira: Scalable hash-grid compression for implicit neural representations. In CVPR, pages 17513–17524, 2023.
- Joint convolutional analysis and synthesis sparse representation for single image layer separation. In ICCV, pages 1708–1716, 2017.
- From sky to the ground: A large-scale benchmark and simple baseline towards real rain removal. In CVPR, pages 12097–12107, 2023.
- Local implicit grid representations for 3d scenes. In CVPR, pages 6001–6010, 2020a.
- Multi-scale progressive fusion network for single image deraining. In CVPR, pages 8346–8355, 2020b.
- Focal frequency loss for image reconstruction and synthesis. In ICCV, pages 13919–13929, 2021.
- Automatic single-image-based rain streaks removal via image decomposition. IEEE TIP, 21(4):1742–1755, 2011.
- Zero-shot blind image denoising via implicit neural representations. arXiv preprint arXiv:2204.02405, 2022a.
- Mssnet: Multi-scale-stage network for single image deblurring. In ECCV, pages 524–539, 2022b.
- Adam: A method for stochastic optimization. In ICLR, 2014.
- Ffcv: Accelerating training by removing data bottlenecks. In CVPR, pages 12011–12020, 2023.
- Local texture estimator for implicit representation function. In CVPR, pages 1929–1938, 2022.
- Deep scale-space mining network for single image deraining. In CVPRW, pages 4276–4285, 2022.
- Recurrent squeeze-and-excitation context aggregation net for single image deraining. In ECCV, pages 254–269, 2018.
- Rain streak removal using layer priors. In CVPR, pages 2736–2744, 2016.
- Scale-aware modulation meet transformer. In ICCV, 2023.
- Sgdr: Stochastic gradient descent with warm restarts. arXiv preprint arXiv:1608.03983, 2016.
- Removing rain from a single image via discriminative sparse coding. In ICCV, pages 3397–3405, 2015.
- Intriguing findings of frequency selection for image deblurring. In AAAI, pages 1905–1913, 2023.
- Pyramid attention network for image restoration. IJCV, pages 1–19, 2023.
- Nerf: Representing scenes as neural radiance fields for view synthesis. Communications of the ACM, 65(1):99–106, 2021.
- Neural image representations for multi-image fusion and layer separation. In ECCV, pages 216–232, 2022.
- Scope of validity of psnr in image/video quality assessment. Electronics Letters, 44(13):800–801, 2008.
- Single image defocus deblurring via implicit neural inverse kernels. In ICCV, pages 12600–12610, 2023.
- Progressive image deraining networks: A better and simpler baseline. In CVPR, pages 3937–3946, 2019.
- Ice: Implicit coordinate encoder for multiple image neural representation. IEEE TIP, 2023.
- Miner: Multiscale implicit neural representation. In ECCV, pages 318–333, 2022.
- Implicit neural representations for image compression. In ECCV, pages 74–91, 2022.
- Joint implicit image function for guided depth super-resolution. In ACM MM, pages 4390–4399, 2021.
- Scale-recurrent network for deep image deblurring. In CVPR, pages 8174–8182, 2018.
- Maxim: Multi-axis mlp for image processing. In CVPR, pages 5769–5780, 2022.
- Attention is all you need. NeurIPS, 30, 2017.
- A model-driven deep neural network for single image rain removal. In CVPR, pages 3103–3112, 2020a.
- Spatial attentive single-image deraining with a high quality real rain dataset. In CVPR, pages 12270–12279, 2019.
- A hierarchical approach for rain or snow removing in a single color image. IEEE TIP, 26(8):3936–3950, 2017.
- Rethinking image deraining via rain streaks and vapors. In ECCV, pages 367–382, 2020b.
- Image quality assessment: from error visibility to structural similarity. IEEE TIP, 13(4):600–612, 2004.
- Uformer: A general u-shaped transformer for image restoration. In CVPR, pages 17683–17693, 2022.
- Image de-raining transformer. IEEE TPAMI, 2022.
- Signal processing for implicit neural representations. NeurIPS, 35:13404–13418, 2022.
- Implicit neural representation for cooperative low-light image enhancement. In ICCV, 2023.
- Deep joint rain detection and removal from a single image. In CVPR, pages 1357–1366, 2017.
- Structure-preserving deraining with residue channel prior guidance. In ICCV, pages 4238–4247, 2021.
- Multi-stage progressive image restoration. In CVPR, pages 14821–14831, 2021.
- Restormer: Efficient transformer for high-resolution image restoration. In CVPR, pages 5728–5739, 2022.
- Density-aware single image de-raining using a multi-stream dense network. In CVPR, pages 695–704, 2018.
- Joint bi-layer optimization for single-image rain streak removal. In ICCV, pages 2526–2534, 2017.