Research map / Image generation / Image restoration
Erasing Appearance Preservation in Optimization-based Smoothing
Lvmin Zhang, Chengze Li, Yi JI, Chunping Liu, Tien-tsin Wong
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 EAP (knapsack-erased appearance preservation) | L0 · Texture decomposition (Cornell Box) | PSNR: 13.45 -> 27.34 | Improved |
| add EAP (knapsack-erased appearance preservation) | L1 · Texture decomposition (Cornell Box) | PSNR: 18.49 -> 28.58 | Improved |
| add EAP | TV · Texture decomposition (Cornell Box) | PSNR: 14.07 -> 22.60 | Improved |
| add EAP | RTV · Texture decomposition (Cornell Box) | PSNR: 16.32 -> 25.20 | Improved |
| add EAP | Bell · IIW | WHDR (%): 20.6 -> 19.9 | Improved |
| add EAP | DL1 · IIW | WHDR (%): 17.7 -> 16.2 | Improved |
| add EAP | DL1 · SAW | AP (%): 93.57 -> 95.21 | Improved |
| raise erasing capacity u from 0.10 to 0.50 (ablation) | EAP + DL1 · IIW | WHDR (%): 16.9 -> 52.5 | Hurt |
| raise erasing capacity u from 0.10 to 0.50 (ablation) | EAP + Bell · SAW | AP (%): 94.43 -> 77.52 | Hurt |
| add EAP | Bell · AMT shadow enhancement | average human rank: 3.55 -> 1.71 | Improved |
| add EAP | DL1 · AMT specular reflection removal | average human rank: 3.47 -> 1.34 | Improved |
| use optimization-based intrinsic methods instead of deep models | CGIntrinsics · AMT shadow enhancement | average human rank: 5.70 -> 3.55 | 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.
- GLoSH — GLoSH: Global-Local Spherical Harmonics for Intrinsic Image Decomposition (ICCV 2019) · better on 2 of 2 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
- Generative Image Layer Decomposition with Visual Effects (CVPR 2025 · similarity 0.84)
- Prune and Repaint: Content-Aware Image Retargeting for any Ratio (NeurIPS 2024 · similarity 0.83)
- GeoRemover: Removing Objects and Their Causal Visual Artifacts (NeurIPS 2025 · similarity 0.83)
- Polarization Guided Mask-Free Shadow Removal (AAAI 2025 · similarity 0.83)
- Move and Act: Enhanced Object Manipulation and Background Integrity for Image Editing (AAAI 2025 · similarity 0.83)
- FIRM: Flexible Interactive Reflection ReMoval (AAAI 2025 · similarity 0.83)
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.
