Research map / 3D vision / Human pose estimation
Dense Intrinsic Appearance Flow for Human Pose Transfer
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 dense intrinsic appearance flow and visibility maps instead of pose/keypoint-only guidance | human pose transfer · DeepFashion test | FashionIS: 3.804 -> 4.898 | Helped |
| use dense intrinsic appearance flow and visibility maps instead of pose/keypoint-only guidance | human pose transfer · DeepFashion test | AttrRec-k (%): 19.017 -> 21.065 | Helped |
| use dense intrinsic appearance flow and visibility maps instead of pose/keypoint-only guidance | human pose transfer · DeepFashion test | AttrRec-k (%): 43.812 -> 49.044 | Helped |
| use two separate encoders for appearance and pose instead of one concatenated-input encoder (ablation) | Dense Intrinsic Appearance Flow · DeepFashion test | FashionIS: 3.927 -> 4.898 | Helped |
| add flow-guided feature warping (ablation) | Dense Intrinsic Appearance Flow · DeepFashion test | FashionIS: 4.119 -> 4.898 | Helped |
| add visibility-map-aware gating (ablation) | Dense Intrinsic Appearance Flow · DeepFashion test | FashionIS: 4.491 -> 4.898 | Helped |
| add pixel-level warping to the feature-warped output (ablation) | Dense Intrinsic Appearance Flow · DeepFashion test | FashionIS: 4.800 -> 4.898 | Helped |
Closest papers by meaning
- Human Appearance Transfer (CVPR 2018 · similarity 0.88)
- Towards Hard-pose Virtual Try-on via 3D-aware Global Correspondence Learning (NeurIPS 2022 · similarity 0.88)
- H-Flow: Self-supervised Human Scene Flow via Physics-inspired Joint Multi-modal Learning (NeurIPS 2026 · similarity 0.85)
- GuideFlow3D: Optimization-Guided Rectified Flow For Appearance Transfer (NeurIPS 2025 · similarity 0.85)
- ODHSR: Online Dense 3D Reconstruction of Humans and Scenes from Monocular Videos (CVPR 2025 · similarity 0.85)
- Flow-NeRF: Joint Learning of Geometry, Poses, and Dense Flow within Unified Neural Representations (CVPR 2025 · 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.