Hint, Not Hole: Localization-Robust Video Overlay Removal
cs.CV
Submitted: 2026-11-10
Updated: 2026-11-11
Project page: https://zdyshine.github.io/HintErase
Terminology
Sources
- DiffuEraser: A Diffusion Model for Video Inpainting
- DeOverlay: A Benchmark and Source-Anchored Flow for Video Overlay Removal
- SEDiT: Mask-Free Video Subtitle Erasure via One-step Diffusion Transformer
- VideoPainter: Any-length Video Inpainting and Editing with Plug-and-Play Context Control
- VACE: All-in-One Video Creation and Editing
- EraserDiT: Fast Video Inpainting with Diffusion Transformer Model
- Gen-L-Video: Multi-Text to Long Video Generation via Temporal Co-Denoising
- Unified Long Video Inpainting and Outpainting via Overlapping High-Order Co-Denoising
- FlashVSR+: Efficient Video Super-Resolution via Hybrid Attention and Reward-Yield Distribution Matching
Related papers
- TEA: Text Encoder Alignment for Robust Concept Erasure in Text-to-Image Models
- Do Egocentric Video-Language Models Capture Both Hand- and Object-Centric Cues?
- Aura: Consistent Multi-Subject Video Generation via VLM-Grounded Semantic Alignment
- PVRF: All-in-one Adverse Weather Removal via Prior-modulated and Velocity-constrained Rectified Flow
- DLWM: Diverse Latent World Models for Efficient Multimodal Reasoning
- Out of Sight, Still in Mind: Token Compression for Omni-LLMs