So Predictable! Continuous 3D Hand Trajectory Prediction in Virtual Reality
summary
Kinematic regression predicts continuous 3D hand trajectories for VR using OptiTrack data from 20 participants and gaming plus pointing tasks, achieving RMSE 0.80 cm at 100 ms, 0.85 cm at 200 ms, 3.15 cm at 300 ms and <1 cm for prediction under 200 ms, with an R2 drop after ~340 ms. The model is user- and activity-independent, runs cheaply, enables early collision/haptic triggers, reconstructs missing tracker segments, and validates on 3 new users and tasks.
AI-written summary · about
related talks
10mProactive Sensing for Improving Hand Pose Estimation
14mBack-Hand-Pose: 3D Hand Pose Estimation for a Wrist-worn Camera via Dorsum Deformation Network
06mWR-Hand: Wearable Armband can Track User's Hand
17mPredicTouch: A System to Reduce Touchscreen Latency using Neural Networks and Inertial ...
19mATK: Enabling Ten-Finger Freehand Typing in Air Based on 3D Hand Tracking Data
10mHAGI: Head-Assisted Gaze Imputation for Mobile Eye Trackers
27mAnalysis of sports scenes using computer vision models
19mBackHand: Sensing Hand Gestures via Back of the Hand