Research map / 3D vision / Human pose estimation
Robust Hand Pose Estimation During the Interaction With an Unknown Object
Chiho Choi, Sang Ho Yoon, Chin-Ning Chen, Karthik Ramani
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 spatial fusion (averaging original and synthesized images) (ablation) | reproduction network with grasp classifier · GUN-71 (33 grasps) | grasp accuracy: 37.75% to 41.00% | Improved |
| add the reproduction network and fusion compared with original images (ablation) | grasp classifier · GUN-71 (33 grasps) | grasp accuracy: 39.75 -> 41.00 | Improved |
| use synthesized images only, without fusion (ablation) | grasp classifier · GUN-71 (33 grasps) | grasp accuracy: 39.75 -> 37.75 | Hurt |
| use the two-stream hand and object network instead of hand-only (ablation) | grasp classifier · Synthetic hand-object dataset | grasp accuracy: 43.87 -> 55.56 | Improved |
| use the two-stream network instead of object-only (ablation) | grasp classifier · Synthetic hand-object dataset | grasp accuracy: 49.12 -> 55.56 | Improved |
| add the object stream to the hand stream (ablation) | orientation classifier · Synthetic hand-object dataset | orientation accuracy: 59.31 -> 60.50 | Improved |
| add the hand stream to the object stream (ablation) | orientation classifier · Synthetic hand-object dataset | orientation accuracy: 51.12 -> 60.50 | Improved |
| use fused data instead of synthesized data (ablation) | pose estimation · Synthetic hand-object dataset | mean joint angle error (degrees): 12.11 -> 10.17 | Improved |
| use fused data instead of original data (ablation) | pose estimation · Synthetic hand-object dataset | mean joint angle error (degrees): 10.73 -> 10.17 | Improved |
| use heatmap regression ConvNet instead of random forest segmentation (ablation) | hand localization · Synthetic hand-object dataset | center error (pixels): 27.6 -> 6.7 | Improved |
| add more convolutional layers to the localization network (ablation) | localization network · Synthetic hand-object dataset | localization accuracy | No clear effect |
Closest papers by meaning
- Depth-Based 3D Hand Pose Estimation: From Current Achievements to Future Goals (CVPR 2018 · similarity 0.93)
- DeepHand: Robust Hand Pose Estimation by Completing a Matrix Imputed With Deep Features (CVPR 2016 · similarity 0.93)
- Collaborative Learning for Hand and Object Reconstruction With Attention-Guided Graph Convolution (CVPR 2022 · similarity 0.92)
- Generalized Feedback Loop for Joint Hand-Object Pose Estimation (TPAMI 2019 · similarity 0.92)
- Real-Time 3D Hand Pose Estimation with 3D Convolutional Neural Networks (TPAMI 2018 · similarity 0.91)
- End-to-End Detection and Pose Estimation of Two Interacting Hands (ICCV 2021 · similarity 0.91)
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.
