Research map / 3D vision / Human pose estimation
MonoCap: Monocular Human Motion Capture using a CNN Coupled with a Geometric Prior
Xiaowei Zhou, Menglong Zhu, Georgios Pavlakos, Spyridon Leonardos, Konstantinos G. Derpanis, Kostas Daniilidis
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 temporal smoothness (ablation) | MonoCap fine-tuned+specific · Human3.6M | per joint error (mm): 79.6 -> 77.8 | Improved |
| add temporal smoothness (ablation) | MonoCap perspective, generic detector · Human3.6M | per joint error (mm): 106.0 -> 102.2 | Improved |
| add EM marginalization of 2D uncertainty (ablation) | MonoCap perspective, fine-tuned · Human3.6M | per joint error (mm): 81.4 -> 79.6 | Improved |
| use EM instead of MAP estimation (ablation) | MonoCap fine-tuned · Human3.6M | per joint error (mm): 79.3 -> 77.8 | Improved |
| use perspective instead of orthographic camera (ablation) | MonoCap fine-tuned · Human3.6M | per joint error (mm): 92.5 -> 77.8 | Improved |
| use perspective instead of orthographic camera | MonoCap given 2D joints · Human3.6M | mean per joint error (mm): 88.7 -> 74.2 | Improved |
| fine-tune 2D detector on Human3.6M | MonoCap nonspecific dictionary · Human3.6M | mean per joint error (mm): 107.8 -> 82.7 | Improved |
| use action-specific pose dictionary | MonoCap fine-tuned detector · Human3.6M | mean per joint error (mm): 82.7 -> 77.8 | Improved |
| use action-specific pose dictionary | MonoCap generic detector · Human3.6M | mean per joint error (mm): 107.8 -> 102.2 | Improved |
| enlarge dictionary from 32 to 128 atoms (ablation) | MonoCap · Human3.6M Directions S9 | reconstruction error (mm): 48.1 -> 44.4 | Improved |
| use perspective model with 2D joints given (reconstruction) | MonoCap · Human3.6M Protocol I | reconstruction error (mm): 51.1 -> 50.5 | No clear effect |
Closest papers by meaning
- Sparseness Meets Deepness: 3D Human Pose Estimation From Monocular Video (CVPR 2016 · similarity 0.92)
- Monocular Total Capture: Posing Face, Body, and Hands in the Wild (CVPR 2019 · similarity 0.91)
- Mocap-2-to-3: Multi-view Lifting for Monocular Motion Recovery with 2D Pretraining (CVPR 2026 · similarity 0.91)
- A Simple yet Effective Baseline for 3D Human Pose Estimation (ICCV 2017 · similarity 0.90)
- Learning to Fuse 2D and 3D Image Cues for Monocular Body Pose Estimation (ICCV 2017 · similarity 0.90)
- Deep Kinematics Analysis for Monocular 3D Human Pose Estimation (CVPR 2020 · similarity 0.90)
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.
