Research map / 3D vision / Depth and stereo
4D Human Body Correspondences From Panoramic Depth Maps
Read the paperCheck 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 panoramic depth maps instead of regular perspective depth maps (ablation) | Ours correspondence matching · FAUST | correspondence accuracy | Helped |
| Use the modified hourglass architecture instead of the compared CNN architecture (ablation) | Ours correspondence matching · FAUST | correspondence accuracy | Helped |
| Add correspondence refinement using temporal smoothness and geodesic consistency (ablation) | Ours correspondence matching · MIT Jumping and Swing | correspondence error | Helped |
| Use the proposed autoencoder instead of PCA-based compression | Ours mesh compression · Jumping, March2, Swing, Yoga, Boxing, and Dance | KG error | Helped |
| Use the proposed correspondence method instead of the previous learning-based approach | Ours correspondence matching · MIT Jumping and Swing | correspondence error | Helped |
Closest papers by meaning
- Dense Human Body Correspondences Using Convolutional Networks (CVPR 2016 · similarity 0.89)
- Neural Descent for Visual 3D Human Pose and Shape (CVPR 2021 · similarity 0.87)
- 3D Human Mesh Regression With Dense Correspondence (CVPR 2020 · similarity 0.87)
- Textured Neural Avatars (CVPR 2019 · similarity 0.87)
- End-to-End Recovery of Human Shape and Pose (CVPR 2018 · similarity 0.86)
- Self-Supervised 3D Human Mesh Recovery from a Single Image with Uncertainty-Aware Learning (AAAI 2024 · similarity 0.85)
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.