Research map / 3D vision / Human pose estimation
Deformer: Dynamic Fusion Transformer for Robust Hand Pose Estimation
Qichen Fu, Xingyu Liu, Ran Xu, Juan Carlos Niebles, Kris M. Kitani
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 Dynamic Fusion Module over frame-wise center prediction (ablation) | Deformer · DexYCB (All) | MPJPE (mm): 5.72 -> 5.22 | Improved |
| learned dynamic confidence instead of simple average of deformed predictions (ablation) | Deformer · DexYCB (All) | MPJPE (mm): 5.40 -> 5.22 | Improved |
| learned confidence instead of occlusion-level weighting (ablation) | Deformer · DexYCB (Occlusion 75%-100%) | MPJPE (mm): 6.33 -> 6.34 | No clear effect |
| add Dynamic Fusion Module on HO3D (ablation) | Deformer · HO3D | Joint error (mm): 9.8 -> 9.4 | Improved |
| replace MSE with maxMSE loss (ablation) | Deformer · DexYCB (All) | MPJPE (mm): 5.63 -> 5.22 | Improved |
| add maxMSE loss with no discriminator (ablation) | Deformer · DexYCB (All) | MPJPE (mm): 5.72 -> 5.43 | Improved |
| add motion discriminator (adversarial training) (ablation) | Deformer · DexYCB (All) | MPJPE (mm): 5.43 -> 5.22 | Improved |
| add motion discriminator without maxMSE (ablation) | Deformer · DexYCB (All) | MPJPE (mm): 5.72 -> 5.63 | Improved |
| Deformer versus best monocular method on HO3D | Deformer · HO3D | Joint error (mm): 9.4 vs 9.1 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- VIBE — VIBE: Video Inference for Human Body Pose and Shape Estimation (CVPR 2020) · better on 2 of 2 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
- Depth-Based Hand Pose Estimation: Methods, Data, and Challenges (IJCV 2018 · similarity 0.90)
- EasyHOI: Unleashing the Power of Large Models for Reconstructing Hand-Object Interactions in the Wild (CVPR 2025 · similarity 0.90)
- What's in Your Hands? 3D Reconstruction of Generic Objects in Hands (CVPR 2022 · similarity 0.90)
- WiLoR: End-to-end 3D Hand Localization and Reconstruction in-the-wild (CVPR 2025 · similarity 0.89)
- Dyn-HaMR: Recovering 4D Interacting Hand Motion from a Dynamic Camera (CVPR 2025 · similarity 0.89)
- UniHOPE: A Unified Approach for Hand-Only and Hand-Object Pose Estimation (CVPR 2025 · similarity 0.89)
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.
