Research map / Video / Video generation
OmniHuman: A Large-scale Dataset and Benchmark for Human-Centric Video Generation
Lei Zhu, xing cai, Yingjie Chen, Li YiHeng, Binxin Yang, Hao Liu, Jie Chen, Chen Li, Jing LYU
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 |
|---|---|---|---|
| fine-tune on 20% of OmniHuman data | LTX-2 · OHBench I2AV | KL divergence (audio): 1.04 -> 0.77 | Improved |
| fine-tune on OmniHuman subset | LTX-2 · OHBench I2AV | AbS: 3.69 -> 4.13 | Improved |
| fine-tune on OmniHuman subset | LTX-2 · OHBench I2AV | T-A: 0.28 -> 0.35 | Improved |
| fine-tune on OmniHuman subset | LTX-2 · OHBench I2AV | DD: 0.601 -> 0.665 | Improved |
| fine-tune on OmniHuman subset | LTX-2 · OHBench I2AV | IC* dual-person ID consistency: 0.562 -> 0.589 | Improved |
| fine-tune on OmniHuman subset | LTX-2 · OHBench I2AV | CN contact naturalness: 6.98 -> 7.32 | Improved |
| fine-tune on OmniHuman subset | LTX-2 · OHBench I2AV | IQ: 0.720 -> 0.721 | No clear effect |
| fine-tune on OmniHuman subset | LTX-2 · OHBench I2AV | LR listener realism: 8.94 -> 8.92 | No clear effect |
| fine-tune on OmniHuman subset | LTX-2 · OHBench I2AV | WER: 0.29 -> 0.27 | Improved |
Closest papers by meaning
- OpenHumanVid: A Large-Scale High-Quality Dataset for Enhancing Human-Centric Video Generation (CVPR 2025 · similarity 0.93)
- Video-Bench: Human-Aligned Video Generation Benchmark (CVPR 2025 · similarity 0.89)
- HumanVid: Demystifying Training Data for Camera-controllable Human Image Animation (NeurIPS 2024 · similarity 0.89)
- HOIGen-1M: A Large-scale Dataset for Human-Object Interaction Video Generation (CVPR 2025 · similarity 0.88)
- MVHumanNet: A Large-scale Dataset of Multi-view Daily Dressing Human Captures (CVPR 2024 · similarity 0.88)
- Auteur: Language-Driven Cinematographic Framing for Human-Centric Video Generation (NeurIPS 2026 · similarity 0.88)
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.
