Research map / Video / Video generation
ExpPortrait: Expressive Portrait Generation via Personalized Representation
Junyi Wang, Yudong Guo, Boyang Guo, Shengming Yang, Juyong Zhang
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 |
|---|---|---|---|
| replace SMPL-X head proxy with the personalized head representation as control signal (ablation) | ExpPortrait DiT generator · portrait reenactment (qualitative) | expressiveness and identity fidelity | Improved |
| add identity-adaptive expression transfer module instead of directly applying expression offsets (ablation) | ExpPortrait · cross-reenactment (qualitative) | facial expressiveness | Improved |
| use ExpPortrait instead of Wan-Animate | portrait video generation · Self-reenactment | SSIM: 0.84360 -> 0.85908 | Improved |
| use ExpPortrait instead of Wan-Animate | portrait video generation · Self-reenactment | PSNR: 26.5109 -> 26.5507 | No clear effect |
| use ExpPortrait instead of Wan-Animate | portrait video generation · Self-reenactment | LPIPS: 0.21397 -> 0.18394 | Improved |
| use ExpPortrait instead of X-NeMo on cross-reenactment expression accuracy | portrait video generation · Cross-reenactment | AED: 0.17092 -> 0.21092 | Hurt |
| use ExpPortrait instead of X-NeMo on cross-reenactment pose accuracy | portrait video generation · Cross-reenactment | APD: 0.02101 -> 0.01277 | Improved |
| use ExpPortrait instead of LivePortrait | portrait video generation · Cross-reenactment | CSIM: 0.72869 -> 0.72917 | Improved |
| use ExpPortrait instead of Follow-Your-Emoji | portrait video generation · Self-reenactment | LPIPS: 0.23645 -> 0.18394 | Improved |
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.
- X-NeMo — X-NeMo: Expressive Neural Motion Reenactment via Disentangled Latent Attention (ICLR 2025) · better on 3 of 4 shared results
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
- RealPortrait: Realistic Portrait Animation with Diffusion Transformers (AAAI 2025 · similarity 0.92)
- DiffPortrait3D: Controllable Diffusion for Zero-Shot Portrait View Synthesis (CVPR 2024 · similarity 0.90)
- Hallo3: Highly Dynamic and Realistic Portrait Image Animation with Video Diffusion Transformer (CVPR 2025 · similarity 0.89)
- X-NeMo: Expressive Neural Motion Reenactment via Disentangled Latent Attention (ICLR 2025 · similarity 0.89)
- GoHD: Gaze-oriented and Highly Disentangled Portrait Animation with Rhythmic Poses and Realistic Expressions (AAAI 2025 · similarity 0.88)
- HunyuanPortrait: Implicit Condition Control for Enhanced Portrait Animation (CVPR 2025 · 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.
