Research map / 3D vision / Human pose estimation
Lie-X: Depth Image Based Articulated Object Pose Estimation, Tracking, and Action Recognition on Lie Groups
Chi Yun Xu, Lakshmi Narasimhan Govindarajan, Yu Zhang, Li Cheng
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 multiple initial poses (ablation) | Lie-X · Fish pose estimation | average joint error (mm): 3.28 -> 0.68 | Improved |
| add multiple initial poses (ablation) | Lie-X · Mouse pose estimation | average joint error (mm): 13.27 -> 6.64 | Improved |
| add multiple initial poses (ablation) | Lie-X · NYU hand pose estimation | average joint error (mm): 20.50 -> 14.51 | Improved |
| add learned internal metric to choose among candidates (ablation) | Lie-X · Fish pose estimation | average joint error (mm): 1.71 -> 0.68 | Improved |
| add learned internal metric to choose among candidates (ablation) | Lie-X · Mouse pose estimation | average joint error (mm): 9.82 -> 6.64 | Improved |
| add learned internal metric to choose among candidates (ablation) | Lie-X · NYU hand pose estimation | average joint error (mm): 16.72 -> 14.51 | Improved |
| use temporal tracking instead of single-frame pose estimation (ablation) | Lie-X · Mouse tracking | average joint error (mm): 8.42 -> 7.19 | Improved |
| use tangent vector (Lie algebra) features instead of joint positions (ablation) | Action recognition · Fish action recognition | average classification accuracy (%): 79.19 -> 91.33 | Improved |
| replace Lie-X with CNN (AlexNet) baseline (ablation) | Pose estimation · NYU hand pose estimation | average joint error (mm): 14.51 -> 18.82 | Hurt |
| vary internal parameters (initial poses, rounds, trees, depth) (ablation) | Lie-X · Fish, mouse, hand pose estimation | average joint error (mm) | No clear effect |
Later papers that compared with this method
3 later accepted papers report Lie-X as a baseline in their results tables; 3 of them report a better number on most of the results they share with it.
- Dual Grid Net: Hand Mesh Vertex Regression from Single Depth Maps (ECCV 2020 · better on 1 of 1 shared result)
- Self-Supervised 3D Hand Pose Estimation Through Training by Fitting (CVPR 2019 · better on 1 of 1 shared result)
- Generalized Feedback Loop for Joint Hand-Object Pose Estimation (TPAMI 2019 · better on 1 of 1 shared result)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Nonparametric Object and Parts Modeling With Lie Group Dynamics (CVPR 2020 · similarity 0.86)
- REACTO: Reconstructing Articulated Objects from a Single Video (CVPR 2024 · similarity 0.85)
- Deep Learning on Lie Groups for Skeleton-Based Action Recognition (CVPR 2017 · similarity 0.85)
- SAOR: Single-View Articulated Object Reconstruction (CVPR 2024 · similarity 0.84)
- Learning Implicit Representation for Reconstructing Articulated Objects (ICLR 2024 · similarity 0.84)
- DIPO: Dual-State Images Controlled Articulated Object Generation Powered by Diverse Data (NeurIPS 2025 · similarity 0.83)
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.
