Research map / 3D vision / Face and shape
InverseFaceNet: Deep Monocular Inverse Face Rendering
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 |
|---|---|---|---|
| add model-space loss (ablation) | AlexNet inverse face rendering · CelebA | photometric error [8 bits]: 46.26 ± 12.42 -> 39.71 ± 9.86 | Helped |
| add model-space loss (ablation) | AlexNet inverse face rendering · CelebA | geometric error [mm]: 2.91 ± 0.99 -> 2.77 ± 1.00 | Helped |
| add model-space loss (ablation) | AlexNet inverse face rendering · CelebA | intersection over union [%]: 90.44 ± 3.81 -> 92.51 ± 2.59 | Helped |
| add self-supervised bootstrapping from real images (ablation) | AlexNet with model-space loss · CelebA | photometric error [8 bits]: 39.71 ± 9.86 -> 34.03 ± 7.56 | Helped |
| add self-supervised bootstrapping from real images (ablation) | AlexNet with model-space loss · CelebA | geometric error [mm]: 2.77 ± 1.00 -> 2.11 ± 0.84 | Helped |
| add self-supervised bootstrapping from real images (ablation) | AlexNet with model-space loss · CelebA | intersection over union [%]: 92.51 ± 2.59 -> 93.96 ± 2.08 | Helped |
| use ResNet-101 instead of AlexNet | inverse face rendering with model-space loss · CelebA | geometric error [mm]: 2.77 ± 1.00 -> 2.54 ± 0.87 | Helped |
| use ResNet-101 instead of AlexNet | inverse face rendering with model-space loss · CelebA | photometric error [8 bits]: 39.71 ± 9.86 -> 41.23 ± 10.58 | Hurt |
| use ResNet-101 instead of AlexNet | inverse face rendering with model-space loss · CelebA | intersection over union [%]: 92.51 ± 2.59 -> 92.07 ± 2.87 | Hurt |
| add self-supervised bootstrapping (ablation) | synthetic-only inverse face rendering · FaceWarehouse | geometric error [mm]: 2.33 mm -> 2.11 mm | Helped |
Closest papers by meaning
- Self-Supervised Multi-Level Face Model Learning for Monocular Reconstruction at Over 250 Hz (CVPR 2018 · similarity 0.90)
- Learning Detailed Face Reconstruction From a Single Image (CVPR 2017 · similarity 0.89)
- Learning Complete 3D Morphable Face Models From Images and Videos (CVPR 2021 · similarity 0.88)
- OSDFace: One-Step Diffusion Model for Face Restoration (CVPR 2025 · similarity 0.88)
- In-N-Out: Faithful 3D GAN Inversion with Volumetric Decomposition for Face Editing (CVPR 2024 · similarity 0.88)
- Synthetic Prior for Few-Shot Drivable Head Avatar Inversion (CVPR 2025 · 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.