Research map / 3D vision / Face and shape
DeepHuman: 3D Human Reconstruction From a Single Image
Zerong Zheng, Tao Yu, Yixuan Wei, Qionghai Dai, Yebin Liu
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 dense semantic map/volume instead of joint heat maps/volumes (ablation) | image-guided vol2vol network · THuman synthetic test set | IoU score (%): 74.16 -> 79.14 | Improved |
| add normal refinement U-Net with volume-to-normal projection (ablation) | DeepHuman · THuman synthetic test set | normal cosine distance: 0.0941 -> 0.0583 | Improved |
| add normal refinement U-Net with volume-to-normal projection (ablation) | DeepHuman · THuman synthetic test set | normal l2-norm error: 0.336 -> 0.262 | Improved |
| use multi-scale volumetric feature transformation instead of finest-scale only (Baseline A) (ablation) | DeepHuman · THuman synthetic data | reconstruction loss / visual quality | Improved |
| use multi-scale VFT instead of coarsest-scale only (Baseline B) or latent code concatenation (Baseline C) (ablation) | DeepHuman · THuman synthetic data | reconstruction quality | Improved |
| use DeepHuman instead of HMR | single-image human reconstruction · THuman synthetic test set | averaged 3D IoU: 41.4% -> 45.7% | Improved |
| use DeepHuman instead of BodyNet | single-image human reconstruction · THuman synthetic test set | averaged 3D IoU: 38.7% -> 45.7% | Improved |
Later papers that compared with this method
8 later accepted papers report DeepHuman as a baseline in their results tables; 6 of them report a better number on most of the results they share with it.
- PaMIR: Parametric Model-Conditioned Implicit Representation for Image-Based Human Reconstruction (TPAMI 2021 · better on 4 of 4 shared results)
- M3D-VTON: A Monocular-to-3D Virtual Try-On Network (ICCV 2021 · better on 3 of 3 shared results)
- Detailed Avatar Recovery From Single Image (TPAMI 2021 · better on 0 of 1 shared result)
- Geo-PIFu: Geometry and Pixel Aligned Implicit Functions for Single-view Human Reconstruction (NeurIPS 2020 · better on 4 of 4 shared results)
- BCNet: Learning Body and Cloth Shape from A Single Image (ECCV 2020 · better on 4 of 4 shared results)
- TexMesh: Reconstructing Detailed Human Texture and Geometry from RGB-D Video (ECCV 2020 · better on 3 of 3 shared results)
- Monocular Real-Time Volumetric Performance Capture (ECCV 2020 · better on 2 of 2 shared results)
- DeepCap: Monocular Human Performance Capture Using Weak Supervision (CVPR 2020 · better on 0 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
- Get3DHuman: Lifting StyleGAN-Human into a 3D Generative Model Using Pixel-Aligned Reconstruction Priors (ICCV 2023 · similarity 0.87)
- Image-to-Voxel Model Translation for 3D Scene Reconstruction and Segmentation (ECCV 2020 · similarity 0.87)
- BodyNet: Volumetric Inference of 3D Human Body Shapes (ECCV 2018 · similarity 0.87)
- SyncHuman: Synchronizing 2D and 3D Generative Models for Single-view Human Reconstruction (NeurIPS 2025 · similarity 0.86)
- DiffHuman: Probabilistic Photorealistic 3D Reconstruction of Humans (CVPR 2024 · similarity 0.86)
- SiTH: Single-view Textured Human Reconstruction with Image-Conditioned Diffusion (CVPR 2024 · similarity 0.86)
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.
