Research map / 3D vision / Face and shape
Photorealistic Facial Texture Inference Using Deep Neural Networks
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 convex combinations of database feature correlations rather than the closest correlation (ablation) | facial texture inference · Texture synthesis evaluation | mesoscopic detail quality | Helped |
| remove the convexity constraints from feature-correlation blending (ablation) | facial texture inference · Texture synthesis evaluation | texture artifacts | Hurt |
| use fewer than all 5 VGG-19 mid-layers (ablation) | multi-scale facial detail representation · Texture synthesis evaluation | texture detail | Hurt |
| replace VGG-19 with CaffeNet (ablation) | facial texture inference · Texture synthesis evaluation | texture artifacts | Hurt |
| replace VGG-19 with VGG-16 (ablation) | facial texture inference · Texture synthesis evaluation | texture artifacts | Hurt |
| use VGG-19 rather than the evaluated shallower networks (ablation) | facial texture inference · Texture synthesis evaluation | texture quality | Helped |
| use convex combinations rather than unconstrained linear combinations (ablation) | facial texture inference · Chicago Face Database user study | real-identical judgments | Helped |
| use unconstrained linear combinations rather than the closest feature correlation (ablation) | facial texture inference · Chicago Face Database user study | real-identical judgments | Helped |
| use closest, unconstrained, or convex feature-correlation variants (ablation) | facial texture inference · Chicago Face Database user study | user-study ratings | No clear effect |
| synthesize textures with the proposed method instead of Light Stage capture | 3D face rendering · Light Stage user study | preferred as more realistic (%): 56% | Mixed |
| replace PCA model fitting with the proposed texture inference | 3D face rendering · Light Stage user study | judged more realistic (%): 72% | Helped |
| use Light Stage capture instead of PCA model fitting | 3D face rendering · Light Stage user study | judged more realistic (%): 74% | Helped |
Closest papers by meaning
- Mesoscopic Facial Geometry Inference Using Deep Neural Networks (CVPR 2018 · similarity 0.91)
- Towards High-Fidelity 3D Face Reconstruction From In-the-Wild Images Using Graph Convolutional Networks (CVPR 2020 · similarity 0.91)
- GANFIT: Generative Adversarial Network Fitting for High Fidelity 3D Face Reconstruction (CVPR 2019 · similarity 0.89)
- A Hierarchical Representation Network for Accurate and Detailed Face Reconstruction From In-the-Wild Images (CVPR 2023 · similarity 0.89)
- FreeUV: Ground-Truth-Free Realistic Facial UV Texture Recovery via Cross-Assembly Inference Strategy (CVPR 2025 · similarity 0.88)
- UV-IDM: Identity-Conditioned Latent Diffusion Model for Face UV-Texture Generation (CVPR 2024 · similarity 0.87)
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.