Research map / Video / Video generation
Shape-Aware Text-Driven Layered Video Editing
Yao-Chih Lee, Ji-Ze Genevieve Jang, Yi-Ting Chen, Elizabeth Qiu, Jia-Bin Huang
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 |
|---|---|---|---|
| Use shape-aware UV deformation and atlas optimization instead of fixed NLA UV mapping | Shape-Aware Text-Driven Layered Video Editing · DAVIS video examples | qualitative shape-aware editing | Improved |
| Add keyframe semantic correspondence while keeping UV mapping fixed (ablation) | Shape-Aware Text-Driven Layered Video Editing · Ablation examples | shape editing | No clear effect |
| Add deformation initialization to deform NLA UV maps using keyframe semantic correspondence (ablation) | Shape-Aware Text-Driven Layered Video Editing · Ablation examples | target-shape restoration | Improved |
| Add atlas optimization to refine the edited atlas and deformation (ablation) | Shape-Aware Text-Driven Layered Video Editing · Ablation examples | visual artifacts | Improved |
| Edit multiple keyframes independently and interpolate remaining frames with FILM | Multi-frame baseline · DAVIS video examples | temporal consistency | Hurt |
| Propagate a single edited keyframe with EbSynth | Single-frame baseline · DAVIS video examples | edit propagation quality | Hurt |
| Use fixed NLA UV mapping for text-driven atlas editing | Text2LIVE · DAVIS video examples | target-shape editing | Hurt |
| Rely on inaccurate NLA mapping for complex motion | Shape-Aware Text-Driven Layered Video Editing · Bear to lion example | visual distortion | Hurt |
Closest papers by meaning
- Pathways on the Image Manifold: Image Editing via Video Generation (CVPR 2025 · similarity 0.86)
- VINCIE-NExT: Unlocking Video Editing from Images via In-Context Modeling (NeurIPS 2026 · similarity 0.86)
- Transforming Image Editors into Video Editors (NeurIPS 2026 · similarity 0.86)
- FLATTEN: optical FLow-guided ATTENtion for consistent text-to-video editing (ICLR 2024 · similarity 0.85)
- TokenFlow: Consistent Diffusion Features for Consistent Video Editing (ICLR 2024 · similarity 0.85)
- CoDeF: Content Deformation Fields for Temporally Consistent Video Processing (CVPR 2024 · 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.
