Research map / Image generation / Image restoration
Under-Display Camera Image Restoration with Scattering Effect
Binbin Song, Xiangyu Chen, Shuning Xu, Jiantao Zhou
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add the scattering branch (ablation) | SRUDC · Synthetic UDC test set | PSNR: 34.66 -> 36.02 | Improved |
| add supervision of scattering parameters m and alpha (ablation) | SRUDC · Synthetic UDC test set | PSNR: 35.13 -> 36.02 | Improved |
| use channel-wise self-attention blocks instead of CNN in scattering branch (ablation) | SRUDC scattering branch · Synthetic UDC test set | PSNR: 35.02 -> 36.02 | Improved |
| add gating operation in GCAB (ablation) | SRUDC image branch · Synthetic UDC test set | PSNR: 35.64 -> 36.02 | Improved |
| add channel attention in GCAB (ablation) | SRUDC image branch · Synthetic UDC test set | PSNR: 35.32 -> 36.02 | Improved |
| use SFT-based feature fusion block instead of concatenation (ablation) | SRUDC · Synthetic UDC test set | PSNR: 35.84 -> 36.02 | Improved |
| use SRUDC-full instead of BNUDC | BNUDC · Synthetic UDC test set | PSNR: 34.91 -> 36.02 | Improved |
| use lightweight SRUDC instead of BNUDC | BNUDC · Synthetic UDC test set | PSNR: 34.91 -> 35.56 | Improved |
| train BNUDC on data from enhanced scattering-aware image formation pipeline | BNUDC · Real-world UDC images | visual quality | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- BNUDC — BNUDC: A Two-Branched Deep Neural Network for Restoring Images From Under-Display Cameras (CVPR 2022) · better on 3 of 3 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Decoupling Degradations with Recurrent Network for Video Restoration in Under-Display Camera (AAAI 2024 · similarity 0.90)
- Removing Diffraction Image Artifacts in Under-Display Camera via Dynamic Skip Connection Network (CVPR 2021 · similarity 0.90)
- Image Restoration for Under-Display Camera (CVPR 2021 · similarity 0.89)
- UDC-SIT: A Real-World Dataset for Under-Display Cameras (NeurIPS 2023 · similarity 0.89)
- Decoupling Scattering: Pseudo-Label Guided NeRF for Scenes with Scattering Media (AAAI 2025 · similarity 0.85)
- SIR-DIFF: Sparse Image Sets Restoration with Multi-View Diffusion Model (CVPR 2025 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
