Research map / 3D vision / Face and shape
JIFF: Jointly-Aligned Implicit Face Function for High Quality Single View Clothed Human Reconstruction
Yukang Cao, Guanying Chen, Kai Han, Wenqi Yang, Kwan-Yee K. Wong
Check 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 jointly aligned 2D image and 3D face-prior features with separate feature-transforming MLPs (ablation) | JIFF face shape reconstruction · Face/head ablation | Face L2 distance: 0.427 -> 0.141 | Improved |
| Add jointly aligned 2D image and 3D face-prior features with separate feature-transforming MLPs (ablation) | JIFF face shape reconstruction · Face/head ablation | Head P2S: 0.761 -> 0.291 | Improved |
| Add jointly aligned 2D image and 3D face-prior features with separate feature-transforming MLPs (ablation) | JIFF face shape reconstruction · Face/head ablation | Head Chamfer: 0.756 -> 0.308 | Improved |
| Add space-aligned 3D features to pixel-aligned 2D features (ablation) | JIFF face shape reconstruction · Face/head ablation | Face L2 distance: 0.427 -> 0.171 | Improved |
| Add space-aligned 3D features to pixel-aligned 2D features (ablation) | JIFF face shape reconstruction · Face/head ablation | Head P2S: 0.761 -> 0.310 | Improved |
| Use dual 3D feature encoders for plain and textured 3DMM inputs instead of a single encoder (ablation) | JIFF face texture reconstruction · Face texture ablation | Face SSIM: 0.7563 -> 0.7649 | Improved |
| Use dual 3D feature encoders for plain and textured 3DMM inputs instead of a single encoder (ablation) | JIFF face texture reconstruction · Face texture ablation | Face LPIPS: 0.1143 -> 0.1107 | Improved |
| Use dual 3D feature encoders for plain and textured 3DMM inputs instead of a single encoder (ablation) | JIFF face texture reconstruction · Face texture ablation | Head L1 error: 0.1137 -> 0.1051 | Improved |
| Use fine-resolution features instead of coarse-resolution features (ablation) | JIFF texture prediction · Face texture ablation | Face SSIM: 0.7126 -> 0.7264 | Improved |
| Add 3D face prior to the fine-resolution texture branch (ablation) | JIFF texture prediction · Face texture ablation | Face SSIM: 0.7264 -> 0.7580 | Improved |
| Use coarse-to-fine prediction with the 3D face prior rather than fine resolution with the 3D prior alone (ablation) | JIFF texture prediction · Face texture ablation | Face SSIM: 0.7580 -> 0.7649 | Improved |
| Extend JIFF with jointly trained face and body branches | JIFF full-body reconstruction · THuman2.0 | Body P2S: 1.747 -> 1.685 | Improved |
Later papers that compared with this method
2 later accepted papers report JIFF as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.
- Implicit Shape and Appearance Priors for Few-Shot Full Head Reconstruction (TPAMI 2025 · better on 1 of 1 shared result)
- G2-MonoDepth: A General Framework of Generalized Depth Inference From Monocular RGB+X Data (TPAMI 2023 · better on 1 of 2 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
- IntegratedPIFu: Integrated Pixel Aligned Implicit Function for Single-View Human Reconstruction (ECCV 2022 · similarity 0.89)
- SIFU: Side-view Conditioned Implicit Function for Real-world Usable Clothed Human Reconstruction (CVPR 2024 · similarity 0.87)
- Fine Structure-Aware Sampling: A New Sampling Training Scheme for Pixel-Aligned Implicit Models in Single-View Human Reconstruction (AAAI 2024 · similarity 0.86)
- HiLo: Detailed and Robust 3D Clothed Human Reconstruction with High-and Low-Frequency Information of Parametric Models (CVPR 2024 · similarity 0.85)
- PSHuman: Photorealistic Single-image 3D Human Reconstruction using Cross-Scale Multiview Diffusion and Explicit Remeshing (CVPR 2025 · similarity 0.85)
- HISR: Hybrid Implicit Surface Representation for Photorealistic 3D Human Reconstruction (AAAI 2024 · similarity 0.85)
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.
