1 paper · 1 filter
Haoran Xie, Atsushi Watatani, Kazunori Miyata
We present an interactive core training system for core training using a monocular camera image as input in this paper. It is commonly expensive to capture human pose using depth c…