Research map / 3D vision / Face and shape
Disentangled3D: Learning a 3D Generative Model With Disentangled Geometry and Appearance From Monocular Images
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 explicit 3D coordinate deformations rather than a 256-dimensional geometry output (ablation) | D3D on FFHQ · FFHQ | Appearance consistency: 0.13 -> 0.05 | Helped |
| Use explicit 3D coordinate deformations rather than a 256-dimensional geometry output (ablation) | D3D on FFHQ · FFHQ | FID: 13.98 -> 28.18 | Hurt |
| Add the inverse deformation network (ablation) | D3D on FFHQ · FFHQ | FID: 19.99 -> 28.18 | Hurt |
| Add pose regularization (ablation) | D3D · VoxCeleb2 | Pose consistency: 0.16 -> 0.03 | Helped |
| Increase the appearance-network capacity relative to the geometry network (ablation) | D3D · FFHQ | Geometry-appearance disentanglement | Hurt |
| Use D3D instead of GRAF | 3D image generation · FFHQ | FID (128×128): 43.32 -> 28.18 | Helped |
| Use D3D instead of GRAF | 3D image generation · VoxCeleb2 | FID (128×128): 35.28 -> 16.51 | Helped |
| Use D3D instead of GRAF | 3D image generation · Cats | FID (128×128): 22.64 -> 16.96 | Helped |
| Use D3D instead of GRAF | 3D image generation · CARLA | FID (128×128): 37.53 -> 31.13 | Helped |
| Add geometry-appearance disentanglement and dense-correspondence constraints | D3D on FFHQ · FFHQ | FID: 13.22 -> 28.18 | Hurt |
Closest papers by meaning
- Visual Object Networks: Image Generation with Disentangled 3D Representations (NeurIPS 2018 · similarity 0.90)
- Disentangled 3D Scene Generation with Layout Learning (ICML 2024 · similarity 0.88)
- What You See is What You GAN: Rendering Every Pixel for High-Fidelity Geometry in 3D GANs (CVPR 2024 · similarity 0.87)
- MVD-Fusion: Single-view 3D via Depth-consistent Multi-view Generation (CVPR 2024 · similarity 0.87)
- Common3D: Self-Supervised Learning of 3D Morphable Models for Common Objects in Neural Feature Space (CVPR 2025 · similarity 0.87)
- Neural Point Cloud Diffusion for Disentangled 3D Shape and Appearance 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.