1 paper
Sanghoon Hong, Hunchul Park, Jonghyuk Park +2
Most of the top-down pose estimation models assume that there exists only one person in a bounding box. However, the assumption is not always correct. In this technical report, we…