Research map / 3D vision / Face and shape
Re-Identification Supervised Texture 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 ReID loss instead of perceptual loss | Human texture generation · Market-1501 test | SSIM: 0.149 -> 0.164 | Helped |
| Use ReID loss instead of pixel-wise ℓ1 loss | Human texture generation · Market-1501 test | SSIM: 0.162 -> 0.164 | Helped |
| Use ReID loss instead of pixel-wise ℓ1 loss | Human texture generation · Market-1501 test | mask-SSIM: 0.374 -> 0.372 | Hurt |
| Use ReID loss instead of perceptual loss | Human texture generation · Market-1501 test | mask-SSIM: 0.356 -> 0.372 | Helped |
| Pretrain on SURREAL++ synthetic data with more diverse textures | Non-local action recognition network · UCF101 | Top-1 Acc. (%): 82.03 -> 86.89 | Helped |
| Pretrain on SURREAL++ rather than SURREAL synthetic data | Non-local action recognition network · UCF101 | Top-1 Acc. (%): 85.83 -> 86.89 | Helped |
| Pretrain on SURREAL++ rather than SURREAL synthetic data | Non-local action recognition network · UCF101 | Top-5 Acc. (%): 96.98 -> 97.04 | Helped |
| Add pose alignment using input-estimated SMPL parameters (ablation) | ReID-supervised human texture generation · Market-1501 test | SSIM: 0.158 -> 0.164 | Helped |
| Add pose alignment using input-estimated SMPL parameters (ablation) | ReID-supervised human texture generation · Market-1501 test | mask-SSIM: 0.365 -> 0.372 | Helped |
| Add body-part features to the ReID feature extractor (ablation) | ReID-supervised human texture generation · Market-1501 test | SSIM: 0.159 -> 0.164 | Helped |
| Add body-part features to the ReID feature extractor (ablation) | ReID-supervised human texture generation · Market-1501 test | mask-SSIM: 0.369 -> 0.372 | Helped |
| Use layer-wise ReID feature loss instead of ℓ1 loss on deep features (ablation) | Human texture generation · Market-1501 test | SSIM: 0.155 -> 0.164 | Helped |
Closest papers by meaning
- SEAS: ShapE-Aligned Supervision for Person Re-Identification (CVPR 2024 · similarity 0.87)
- HumanRef: Single Image to 3D Human Generation via Reference-Guided Diffusion (CVPR 2024 · similarity 0.86)
- From Poses to Identity: Training-Free Person Re-Identification via Feature Centralization (CVPR 2025 · similarity 0.85)
- Alice Benchmarks: Connecting Real World Re-Identification with the Synthetic (ICLR 2024 · similarity 0.84)
- Distilling CLIP with Dual Guidance for Learning Discriminative Human Body Shape Representation (CVPR 2024 · similarity 0.84)
- SCULPT: Shape-Conditioned Unpaired Learning of Pose-dependent Clothed and Textured Human Meshes (CVPR 2024 · similarity 0.84)
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.