Research map / Video / Video generation
Object-WIPER: Training-Free Object and Associated Effect Removal in Videos
Saksham Singh Kushwaha, Sayan Nag, Yapeng Tian, Kuldeep Kulkarni
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 attention scaling (ablation) | Object-WIPER · DAVIS | BG-PSNR: 21.92 -> 23.02 | Improved |
| add attention scaling (ablation) | Object-WIPER · DAVIS | TokSim: 32.97 -> 32.80 | No clear effect |
| add attention scaling (ablation) | Object-WIPER · DAVIS | Text-align: 26.42 -> 26.63 | Improved |
| add timestep-adaptive masking (ablation) | Object-WIPER · DAVIS | TokSim: 32.10 -> 32.80 | Improved |
| add timestep-adaptive masking (ablation) | Object-WIPER · DAVIS | BG-PSNR: 22.73 -> 23.02 | Improved |
| add Gaussian re-initialization of masked tokens (ablation) | Object-WIPER · DAVIS | TokSim: 30.36 -> 32.80 | Improved |
| add Gaussian re-initialization of masked tokens (ablation) | Object-WIPER · DAVIS | BG-PSNR: 23.47 -> 23.02 | Hurt |
| add associated-effect mask MAE (ablation) | Object-WIPER · DAVIS | TokSim: 32.18 -> 32.80 | Improved |
| add associated-effect mask MAE (ablation) | Object-WIPER · DAVIS | Text-align: 26.17 -> 26.63 | Improved |
| use training-free Object-WIPER instead of training-based ROSE | Object-WIPER · WIPER-Bench | TokSim: 30.02 -> 33.09 | Improved |
| use training-free Object-WIPER instead of Attentive-Eraser | Object-WIPER · DAVIS | BG-PSNR: 28.07 -> 23.02 | Hurt |
| use training-free Object-WIPER instead of Propainter | Object-WIPER · WIPER-Bench | BG-PSNR: 41.07 -> 27.53 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 2 on most shared results.
- ROSE — ROSE: Remove Objects with Side Effects in Videos (NeurIPS 2025) · better on 2 of 4 shared results
- Propainter — ProPainter: Improving Propagation and Transformer for Video Inpainting (ICCV 2023) · better on 2 of 4 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
- MiniMax-Remover: Taming Bad Noise Helps Video Object Removal (NeurIPS 2025 · similarity 0.89)
- ROSE: Remove Objects with Side Effects in Videos (NeurIPS 2025 · similarity 0.88)
- SmartEraser: Remove Anything from Images using Masked-Region Guidance (CVPR 2025 · similarity 0.87)
- Generative Omnimatte: Learning to Decompose Video into Layers (CVPR 2025 · similarity 0.87)
- Attentive Eraser: Unleashing Diffusion Model’s Object Removal Potential via Self-Attention Redirection Guidance (AAAI 2025 · similarity 0.86)
- RORem: Training a Robust Object Remover with Human-in-the-Loop (CVPR 2025 · similarity 0.86)
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.
