Research map / 3D vision / Face and shape
BodyGAN: General-Purpose Controllable Neural Human Body Generation
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 BodyGAN's multi-factor condition maps and generator instead of Pix2PixHD* | human image synthesis · Our Test Set | SSIM: 0.7884 -> 0.8307 | Helped |
| Use BodyGAN's multi-factor condition maps and generator instead of Pix2PixHD* | human image synthesis · DeepFashion | FID: 29.7461 -> 6.1654 | Helped |
| Use BodyGAN's multi-factor condition maps and generator instead of Pix2PixHD* | human image synthesis · VITON | LPIPS: 0.0331 -> 0.0146 | Helped |
| Use all proposed pose and appearance condition maps rather than csg only (ablation) | BodyGAN · Our Test Set | SSIM: 0.7466 -> 0.8307 | Helped |
| Add ckp to csg (ablation) | BodyGAN · Our Test Set | SSIM: 0.7466 -> 0.7602 | Helped |
| Add ckp to csg (ablation) | BodyGAN · Our Test Set | LPIPS: 0.0473 -> 0.0477 | Hurt |
| Use cp instead of csg + ckp (ablation) | BodyGAN · Our Test Set | SSIM: 0.7602 -> 0.7694 | Helped |
| Add chand to cp (ablation) | BodyGAN · Our Test Set | SSIM: 0.7694 -> 0.7807 | Helped |
| Add chand to cp (ablation) | BodyGAN · Our Test Set | FID: 5.2623 -> 5.5167 | Hurt |
| Use the full appearance maps ca with cp instead of chand alone (ablation) | BodyGAN · Our Test Set | SSIM: 0.7807 -> 0.8307 | Helped |
| Use the full appearance maps ca with cp instead of chand alone (ablation) | BodyGAN · Our Test Set | FID: 5.5167 -> 3.6760 | Helped |
| Use the full appearance maps ca with cp instead of chand alone (ablation) | BodyGAN · Our Test Set | LPIPS: 0.0445 -> 0.0297 | Helped |
Closest papers by meaning
- InsetGAN for Full-Body Image Generation (CVPR 2022 · similarity 0.91)
- Soft-Gated Warping-GAN for Pose-Guided Person Image Synthesis (NeurIPS 2018 · similarity 0.88)
- Controllable Person Image Synthesis With Attribute-Decomposed GAN (CVPR 2020 · similarity 0.88)
- SCULPT: Shape-Conditioned Unpaired Learning of Pose-dependent Clothed and Textured Human Meshes (CVPR 2024 · similarity 0.86)
- 3DHumanEdit: Multi-modal Body Part-aware Conditioning Information Integration for 3D Human Manipulation (AAAI 2025 · similarity 0.86)
- RealisHuman: A Two-Stage Approach for Refining Malformed Human Parts in Generated Images (AAAI 2025 · 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.