Research map / 3D vision / Face and shape
Single View Metrology in the Wild
Rui Zhu, Xingyi Yang, Yannick Hold-Geoffroy, Federico Perazzi, Jonathan Eisenmann, Kalyan Sunkavalli, Manmohan Chandraker
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| add two cascade refinement layers (SN-L1 to SN-L3) (ablation) | SN-L1 · COCO-Scale-person test | Lvt reprojection error (mean): 0.1502 -> 0.0717 | Improved |
| add two cascade refinement layers (SN-L1 to SN-L3) (ablation) | SN-L1 · KITTI-person | Lvt reprojection error (mean): 0.0980 -> 0.0331 | Improved |
| train with multiple categories (people and cars) (ablation) | SN-L3 · COCO-Scale-person test | Lvt reprojection error (mean): 0.0717 -> 0.0613 | Improved |
| train with multiple categories (people and cars) (ablation) | SN-L3 · KITTI-person | Lvt reprojection error (mean): 0.0331 -> 0.0283 | Improved |
| train with multiple categories (people and cars) (ablation) | SN-L3 · KITTI-person | Ehobj object height error (mean): 0.0956 -> 0.0849 | Improved |
| predict keypoints and upright height with upright ratio (ablation) | SN-L3 · COCO-Scale-person test | Lvt reprojection error (mean): 0.0717 -> 0.0540 | Improved |
| predict keypoints and upright height with upright ratio (ablation) | SN-L3 · COCO-Scale-person A/B realism | user preference: 42.8% -> 57.2% | Improved |
| predict keypoints and upright height with upright ratio (ablation) | SN-L3 · IMDB-23K (non-standing persons) | Ehobj upright height error (mean): 0.1591 -> 0.1212 | Improved |
| predict keypoints and upright height with upright ratio (ablation) | SN-L3 · IMDB-23K | Ehobj upright height error (mean): 0.0891 -> 0.1003 | Hurt |
| predict keypoints and upright height with upright ratio (ablation) | SN-L3 · KITTI-person | Ehobj object height error (mean): 0.0956 -> 0.1014 | Hurt |
| replace PGM baseline with learned SN-L3-kps-can | PGM · COCO-Scale-person A/B realism | user preference: 40.5% -> 59.5% | Improved |
| learn object heights instead of PGM | PGM · IMDB-23K | Ehobj upright height error (mean): 0.2234 -> 0.0891 | Mixed |
Closest papers by meaning
- Fostering Generalization in Single-View 3D Reconstruction by Learning a Hierarchy of Local and Global Shape Priors (CVPR 2021 · similarity 0.88)
- Know Your Neighbors: Improving Single-View Reconstruction via Spatial Vision-Language Reasoning (CVPR 2024 · similarity 0.87)
- Zero-Shot Scene Reconstruction from Single Images with Deep Prior Assembly (NeurIPS 2024 · similarity 0.86)
- 3D Reconstruction with Generalizable Neural Fields using Scene Priors (ICLR 2024 · similarity 0.86)
- LEAP: Liberate Sparse-View 3D Modeling from Camera Poses (ICLR 2024 · similarity 0.85)
- Depth Anything: Unleashing the Power of Large-Scale Unlabeled Data (CVPR 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.
