Research map / 3D vision / Face and shape
Mesoscopic Facial Geometry Inference Using Deep Neural Networks
Read the paperCheck 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 |
|---|---|---|---|
| Separate displacement targets into medium- and high-frequency components (ablation) | Texture-to-Displacement Network · Facial geometry reconstruction | reconstruction quality | Helped |
| Add patch-based super-resolution to increase displacement maps from 1K to 4K (ablation) | Mesoscopic Facial Geometry Inference · Facial geometry reconstruction | faithfulness to mesoscopic facial structures | Helped |
| Use measured photometric-stereo geometry rather than a dark-is-deep heuristic | Facial geometry reconstruction · Facial detail comparison with ground truth | L1 reconstruction error | Helped |
| Use synthesized geometry instead of Light Stage geometry | Mesoscopic Facial Geometry Inference · User study, 6 subjects | most realistic selections (%): 20.7% vs 67.2% | Hurt |
| Use synthesized geometry instead of dark-is-deep geometry | Mesoscopic Facial Geometry Inference · User study, 6 subjects | most realistic selections (%): 20.7% vs 12.1% | Helped |
Closest papers by meaning
- Photorealistic Facial Texture Inference Using Deep Neural Networks (CVPR 2017 · similarity 0.91)
- Learning Detailed Face Reconstruction From a Single Image (CVPR 2017 · similarity 0.87)
- Towards High-Fidelity 3D Face Reconstruction From In-the-Wild Images Using Graph Convolutional Networks (CVPR 2020 · similarity 0.87)
- Modeling Facial Geometry Using Compositional VAEs (CVPR 2018 · similarity 0.87)
- Self-Supervised Multi-Level Face Model Learning for Monocular Reconstruction at Over 250 Hz (CVPR 2018 · similarity 0.87)
- Learning to Decouple the Lights for 3D Face Texture Modeling (NeurIPS 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.