Research map / 3D vision / Human pose estimation
Hand-Transformer: Non-Autoregressive Structured Modeling for 3D Hand Pose Estimation
Lin Huang, Jianchao Tan, Ji Liu, Junsong Yuan
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 |
|---|---|---|---|
| use non-autoregressive structured decoding instead of autoregressive decoding (ablation) | NARHT · ICVL | mean joint error (mm, lower better): 8.57 -> 6.47 | Improved |
| use joint-wise features instead of coordinates as reference pose (ablation) | NARHT · ICVL | mean joint error (mm, lower better): 6.67 -> 6.47 | Improved |
| add self-attention PointNet encoder instead of Point-to-Point encoder (ablation) | NARHT · ICVL | mean joint error (mm, lower better): 6.70 -> 6.47 | Improved |
| add self-attention PointNet encoder instead of Point-to-Point encoder (ablation) | NARHT · ICVL encoder runtime | runtime per frame (ms): 35.90 ms -> 3.80 ms | Improved |
| use non-autoregressive decoding instead of autoregressive Transformer (ablation) | NARHT · Inference speed | FPS: ART much slower; NARHT 43.2 FPS | Improved |
| use NARHT instead of V2V-PoseNet | NARHT · ICVL | mean joint error (mm, lower better): 6.28 vs 6.47 mean error; 3.5 -> 43.2 FPS | Mixed |
| use NARHT instead of DenseReg | NARHT · MSRA | mean joint error (mm, lower better): 7.23 vs 7.55 mean error; 27.8 -> 43.2 FPS | Mixed |
| use NARHT instead of V2V-PoseNet | NARHT · NYU | mean joint error (mm, lower better): 8.42 vs 9.80 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 3 on most shared results.
- V2V-PoseNet — V2V-PoseNet: Voxel-to-Voxel Prediction Network for Accurate 3D Hand and Human Pose Estimation From a Single Depth Map (CVPR 2018) · better on 2 of 4 shared results
- CrossInfoNet — CrossInfoNet: Multi-Task Information Sharing Based Hand Pose Estimation (CVPR 2019) · better on 3 of 3 shared results
- Hand PointNet — Hand PointNet: 3D Hand Pose Estimation Using Point Sets (CVPR 2018) · better on 3 of 3 shared results
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
- 3D Hand Pose Estimation via Articulated Anchor-to-Joint 3D Local Regressors (TPAMI 2025 · similarity 0.91)
- Prior-aware Dynamic Temporal Modeling Framework for Sequential 3D Hand Pose Estimation (ICCV 2025 · similarity 0.90)
- A2J-Transformer: Anchor-to-Joint Transformer Network for 3D Interacting Hand Pose Estimation From a Single RGB Image (CVPR 2023 · similarity 0.90)
- Analyzing the Synthetic-to-Real Domain Gap in 3D Hand Pose Estimation (CVPR 2025 · similarity 0.88)
- Reconstructing Hands in 3D with Transformers (CVPR 2024 · similarity 0.88)
- HandBooster: Boosting 3D Hand-Mesh Reconstruction by Conditional Synthesis and Sampling of Hand-Object Interactions (CVPR 2024 · similarity 0.88)
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.
