Research map / Video / Video generation
VOID: Video Object and Interaction Deletion
Saman Motamed, William Harvey, Benjamin Klein, Luc Van Gool, Zhuoning Yuan, Ta-Ying Cheng
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 |
|---|---|---|---|
| counterfactual physics-aware removal (VOID) instead of Runway text editing | video object removal · Real-world videos, human preference | win %: 18.4 -> 64.8 | Improved |
| VOID versus best baselines on VLM judges | video object removal · Real-world videos, three VLM judges | total score: 25.05 -> 26.13 (Gemini); 20.21 -> 23.16 (GPT-5.2) | Improved |
| VOID versus baselines on synthetic counterfactual benchmark | video object removal · Synthetic benchmark | FVD: 437.88 -> 260.31 | Improved |
| VOID versus ProPainter on LPIPS | video object removal · Synthetic benchmark | LPIPS: 0.0981 -> 0.1156 | Hurt |
| mix Kubric and HUMOTO data instead of a single dataset (size fixed at 1200) (ablation) | VOID training · Real-world videos, Gemini judge | total score: 20.36 / 20.12 -> 21.93 | Improved |
| use VLM-guided quadmask instead of simple trimask (ablation) | VOID · Real-world videos, Gemini judge | total score: 23.39 -> 26.12 | Improved |
| use Gemini 3 Pro as mask VLM rather than Qwen3-32B or GPT 5.2 (ablation) | VOID inference · Real-world videos, Gemini judge | total score: 23.91 / 24.34 -> 26.12 | Improved |
| add flow-warped noise second pass (ablation) | VOID · 10 flagged real-world videos | total score: 23.5 -> 26.0 | Improved |
| second pass on preservation score (ablation) | VOID · 10 flagged real-world videos | Preservation (0-5): 4.90 -> 4.80 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- ProPainter — ProPainter: Improving Propagation and Transformer for Video Inpainting (ICCV 2023) · better on 2 of 3 shared results
- MiniMax-Remover — MiniMax-Remover: Taming Bad Noise Helps Video Object Removal (NeurIPS 2025) · 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
- ROSE: Remove Objects with Side Effects in Videos (NeurIPS 2025 · similarity 0.86)
- Generative Omnimatte: Learning to Decompose Video into Layers (CVPR 2025 · similarity 0.86)
- GeoRemover: Removing Objects and Their Causal Visual Artifacts (NeurIPS 2025 · similarity 0.86)
- Paint by Inpaint: Learning to Add Image Objects by Removing Them First (CVPR 2025 · similarity 0.85)
- Learning Action and Reasoning-Centric Image Editing from Videos and Simulation (NeurIPS 2024 · similarity 0.85)
- MiniMax-Remover: Taming Bad Noise Helps Video Object Removal (NeurIPS 2025 · similarity 0.85)
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.
