Research map / 3D vision / Human pose estimation
PressureVision: Estimating Hand Pressure from a Single RGB Image
Patrick Grady, Chengcheng Tang, Samarth Brahmbhatt, Christopher D. Twigg, Chengde Wan, James Hays, Charles C. Kemp
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 PressureVisionNet instead of 3D pose baseline | hand pressure estimation · PressureVisionDB test (6 unseen participants) | contact IoU: 13.0% -> 55.8% | Improved |
| use PressureVisionNet instead of 3D pose baseline | hand pressure estimation · PressureVisionDB test (6 unseen participants) | temporal accuracy: 78.1% -> 96.2% | Improved |
| add temporal input (4 frames) (ablation) | PressureVisionNet · PressureVisionDB test (6 unseen participants) | volumetric IoU: 41.3% -> 43.1% | Improved |
| add temporal input (4 frames) (ablation) | PressureVisionNet · PressureVisionDB test (6 unseen participants) | MAE: 39.9 Pa -> 39.4 Pa | Improved |
| use RGB color instead of monochrome (ablation) | PressureVisionNet · PressureVisionDB test (6 unseen participants) | volumetric IoU: 37.7% -> 41.3% | Improved |
| use RGB color instead of monochrome (ablation) | PressureVisionNet · PressureVisionDB test (6 unseen participants) | contact IoU: 53.2% -> 55.8% | Improved |
| increase input resolution from 15x12 to 480x384 (ablation) | PressureVisionNet · PressureVisionDB test (6 unseen participants) | volumetric IoU: 11.7% -> 41.3% | Improved |
| increase input resolution from 120x96 to 480x384 (ablation) | PressureVisionNet · PressureVisionDB test (6 unseen participants) | volumetric IoU: 39.5% -> 41.3% | Mixed |
| increase input resolution from 240x192 to 480x384 (ablation) | PressureVisionNet · PressureVisionDB test (6 unseen participants) | volumetric IoU: 39.3% -> 41.3% | Improved |
| classification over 9 log bins instead of direct regression (ablation) | PressureVisionNet · PressureVisionDB test (6 unseen participants) | pressure accuracy | Improved |
| dark vs light skin tone | PressureVisionNet · PressureVisionDB skin tone (5-fold CV) | volumetric IoU: 38.5% -> 39.0% | No clear effect |
| palm press instead of single-finger press | PressureVisionNet · PressureVisionDB test (6 unseen participants) | volumetric IoU: 59.4% -> 33.2% | Hurt |
Closest papers by meaning
- HOPE: Hand-Object Pressure Estimation from Monocular Videos (NeurIPS 2026 · similarity 0.91)
- EgoPressure: A Dataset for Hand Pressure and Pose Estimation in Egocentric Vision (CVPR 2025 · similarity 0.90)
- Posture-Informed Muscular Force Learning for Robust Hand Pressure Estimation (NeurIPS 2024 · similarity 0.88)
- MotionPRO: Exploring the Role of Pressure in Human MoCap and Beyond (CVPR 2025 · similarity 0.87)
- ContactPose: A Dataset of Grasps with Object Contact and Hand Pose (ECCV 2020 · similarity 0.87)
- BodyMAP - Jointly Predicting Body Mesh and 3D Applied Pressure Map for People in Bed (CVPR 2024 · similarity 0.85)
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.
