Event

HKU team proposes YOCO hand motion capture calibration method

1 reports · 1 sources

Researchers from the University of Hong Kong, Yisheng Technology, and Shenzhen Hetao College have proposed YOCO, a few-shot session calibration method for 3D keypoints, used for bias correction of low-cost motion capture gloves in dexterous hand teleoperation; the paper has been accepted by CoRL 2026. YOCO is suited to settings where within-session bias is relatively stable; recalibration may be needed when the glove is re-donned or the bias changes noticeably. The physical experiments were completed by two operators, and a retargeting model was trained separately for each calibration condition, so the 90% reflects the performance of the entire calibration and retargeting pipeline, and it is not yet possible to attribute all of the improvement solely to the calibration module; the relevant measurements include only the pre-deployment computation process and do not include the time for operators to collect calibration motions.