Research map / Video / Video generation
LivePhoto: Real Image Animation with Text-guided Motion Control
Xi Chen, Zhiheng Liu, Mengting Chen, Yutong Feng, Yu Liu, Yujun Shen, Hengshuang Zhao
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 DINOv2 content encoder to reference-latent concatenation (ablation) | LivePhoto baseline · WebVid val frame consistency | DINO score: 82.3 -> 85.9 | Improved |
| add prior noise inversion at inference (ablation) | LivePhoto baseline · WebVid val frame consistency | DINO score: 85.9 -> 90.8 | Improved |
| add motion intensity guidance (ablation) | LivePhoto · WebVid val frame consistency | DINO score: 90.3 -> 90.8 | Improved |
| add text re-weighting (ablation) | LivePhoto · WebVid val frame consistency | CLIP score: 93.9 -> 95.2 | Improved |
| add text re-weighting (ablation) | LivePhoto · User study | Text consistency: 3.3 -> 4.7 | Improved |
| add motion intensity guidance (ablation) | LivePhoto · User study | Text consistency: 2.5 -> 4.7 | Improved |
| LivePhoto vs GEN-2 and Pikalabs on image consistency | LivePhoto · User study | Image consistency: 3.9 -> 3.6 (vs Pikalabs) | Hurt |
| LivePhoto vs GEN-2 on content quality | LivePhoto · User study | Content quality: 4.8 -> 3.7 | Hurt |
| use motion intensity level 2 or level 10 instead of default 5 (ablation) | LivePhoto · qualitative | motion amount | Mixed |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- VideoComposer — VideoComposer: Compositional Video Synthesis with Motion Controllability (NeurIPS 2023) · better on 1 of 1 shared result
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
- Enhancing Motion in Text-to-Video Generation with Decomposed Encoding and Conditioning (NeurIPS 2024 · similarity 0.88)
- MotiF: Making Text Count in Image Animation with Motion Focal Loss (CVPR 2025 · similarity 0.88)
- Enhanced Fine-Grained Motion Diffusion for Text-Driven Human Motion Synthesis (AAAI 2024 · similarity 0.87)
- MotionBooth: Motion-Aware Customized Text-to-Video Generation (NeurIPS 2024 · similarity 0.87)
- Motion Prompting: Controlling Video Generation with Motion Trajectories (CVPR 2025 · similarity 0.87)
- TIV-Diffusion: Towards Object-Centric Movement for Text-driven Image to Video Generation (AAAI 2025 · similarity 0.87)
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.
