Research map / 3D vision / Face and shape
Inverting Generative Adversarial Renderer for Face Reconstruction
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 |
|---|---|---|---|
| add renderer-inverting initialization (ablation) | GAR face reconstruction · Florence Cooperative | point-to-plane distance: 3.20 -> 0.94 | Helped |
| add renderer-inverting initialization (ablation) | GAR face reconstruction · Florence Indoor | point-to-plane distance: 3.21 -> 0.92 | Helped |
| add renderer-inverting initialization (ablation) | GAR face reconstruction · Florence Outdoor | point-to-plane distance: 2.98 -> 0.90 | Helped |
| add normal consistency loss (ablation) | GAR face reconstruction · Florence Cooperative | point-to-plane distance: 1.20 -> 0.94 | Helped |
| add normal consistency loss (ablation) | GAR face reconstruction · Florence Indoor | point-to-plane distance: 1.10 -> 0.92 | Helped |
| add normal consistency loss (ablation) | GAR face reconstruction · Florence Outdoor | point-to-plane distance: 1.40 -> 0.90 | Helped |
| condition the renderer on face normal maps rather than 3DMM parameters or depth (ablation) | GAR · GAR training and optimization | loss convergence | Helped |
| use Normal Injection Module rather than simple normal-map concatenation (ablation) | GAR · GAR training | loss value | Helped |
| use GAR rather than StyleGAN | face image generation · CelebA | FID: 5.17 -> 5.48 | Hurt |
| use GAR rather than StyleGAN | face image generation · FFHQ | FID: 4.40 -> 5.09 | Hurt |
Closest papers by meaning
- GANFIT: Generative Adversarial Network Fitting for High Fidelity 3D Face Reconstruction (CVPR 2019 · similarity 0.90)
- In-N-Out: Faithful 3D GAN Inversion with Volumetric Decomposition for Face Editing (CVPR 2024 · similarity 0.87)
- OFER: Occluded Face Expression Reconstruction (CVPR 2025 · similarity 0.87)
- Dual Encoder GAN Inversion for High-Fidelity 3D Head Reconstruction from Single Images (NeurIPS 2024 · similarity 0.86)
- 3D Facial Expressions through Analysis-by-Neural-Synthesis (CVPR 2024 · similarity 0.86)
- Diffusion-driven GAN Inversion for Multi-Modal Face Image Generation (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.