Research map / Video / Video generation
In-Context Sync-LoRA for Portrait Video Editing
Sagi Polaczek, Or Patashnik, Ali Mahdavi-Amiri, Danny Cohen-Or
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 speech cue to data filtering (ablation) | Sync-LoRA · Portrait video editing benchmark (166 videos) | Speech Corr.: 0.53 -> 0.72 | Improved |
| add pose cue to data filtering (ablation) | Sync-LoRA · Portrait video editing benchmark (166 videos) | Gaze Corr.: 0.67 -> 0.75 | Improved |
| add gaze cue to data filtering (ablation) | Sync-LoRA · Portrait video editing benchmark (166 videos) | Gaze Corr.: 0.67 -> 0.75 | Improved |
| add blink cue to data filtering (ablation) | Sync-LoRA · Portrait video editing benchmark (166 videos) | Blink Corr.: 0.49 -> 0.55 | Improved |
| add pose cue to data filtering (ablation) | Sync-LoRA · Portrait video editing benchmark (166 videos) | Pose Corr.: 0.46 -> 0.55 | Improved |
| add edited pairs to identity-only training data (ablation) | Sync-LoRA · Portrait video editing benchmark (166 videos) | Directional CLIP (image): 0.05 -> 0.57 | Improved |
| add sync filtering and identical pairs to Edit-Only data (ablation) | Sync-LoRA · Portrait video editing benchmark (166 videos) | Speech Corr.: 0.58 -> 0.72 | Improved |
| add sync filtering to random unfiltered pairs (ablation) | Sync-LoRA · Portrait video editing benchmark (166 videos) | Speech Corr.: 0.56 -> 0.72 | Improved |
| add sync filtering to random unfiltered pairs (ablation) | Sync-LoRA · Portrait video editing benchmark (166 videos) | ArcFace Sim.: 0.72 -> 0.75 | Improved |
| add edited pairs to identity-only data (ablation) | Sync-LoRA · Portrait video editing benchmark (166 videos) | Speech Corr.: 0.80 -> 0.72 | Hurt |
| Sync-LoRA vs LucyEdit speech synchronization | Sync-LoRA · Portrait video editing benchmark (166 videos) | Speech Corr.: 0.80 vs 0.72 | Hurt |
| Sync-LoRA vs LucyEdit edit fidelity | Sync-LoRA · Portrait video editing benchmark (166 videos) | Directional CLIP (text-dual): 0.11 -> 0.21 | Improved |
Closest papers by meaning
- Controllable First-Frame-Guided Video Editing via Mask-Aware LoRA Fine-Tuning (ICLR 2026 · similarity 0.87)
- Beyond Wide-Angle Images: Structure-to-Detail Video Portrait Correction via Unsupervised Spatiotemporal Adaptation (AAAI 2026 · similarity 0.87)
- EchoShot: Multi-Shot Portrait Video Generation (NeurIPS 2025 · similarity 0.86)
- VINCIE-NExT: Unlocking Video Editing from Images via In-Context Modeling (NeurIPS 2026 · similarity 0.86)
- HyperLoRA: Parameter-Efficient Adaptive Generation for Portrait Synthesis (CVPR 2025 · similarity 0.85)
- MotionEditor: Editing Video Motion via Content-Aware Diffusion (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.
