Research map / 3D vision / Face and shape
Neural Re-Rendering of Humans from a Single Image
Kripasindhu Sarkar, Dushyant Mehta, Weipeng Xu, Vladislav Golyanik, Christian Theobalt
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 neural UV texture re-rendering instead of CBI | pose transfer · DeepFashion | LPIPS: 0.178 -> 0.164 | Improved |
| use neural UV texture re-rendering instead of CBI | pose transfer · DeepFashion | SSIM: 0.766 -> 0.768 | No clear effect |
| use neural UV texture re-rendering instead of DPT | pose transfer · DeepFashion | LPIPS: 0.206 -> 0.164 | Improved |
| use our method instead of CBI in user study | pose transfer · User study identity | preference rate: 46.2% -> 53.8% | Improved |
| use our method instead of CBI in user study | pose transfer · User study realism | preference rate: 30.8% -> 69.2% | Improved |
| add learned full neural texture instead of inpainted RGB texture (IP baseline) (ablation) | re-rendering pipeline · DeepFashion qualitative | visual quality | Improved |
| add learned full neural texture instead of warping partial texture (Warp baseline) (ablation) | re-rendering pipeline · DeepFashion qualitative | visual quality | Improved |
| condition generator on target DensePose in addition to warped partial texture (WarpCond) (ablation) | Warp baseline · DeepFashion qualitative | visual quality | Improved |
| add intermediate texture loss (ablation) | neural texture pipeline · DeepFashion | convergence speed / final quality | No clear effect |
Closest papers by meaning
- Pose-Guided Human Animation From a Single Image in the Wild (CVPR 2021 · similarity 0.90)
- Re-Identification Supervised Texture Generation (CVPR 2019 · similarity 0.89)
- Human Appearance Transfer (CVPR 2018 · similarity 0.88)
- Combining Implicit Function Learning and Parametric Models for 3D Human Reconstruction (ECCV 2020 · similarity 0.88)
- Refining 3D Human Texture Estimation From a Single Image (TPAMI 2024 · similarity 0.88)
- Textured Neural Avatars (CVPR 2019 · 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.
