weixin_39764569的博客

Ground truth data is attached in the given dataset, containing bounding boxes over each player and all 3 referees (taking the minimum visible X and Y coordinates of each individual) in every single frame (when visible); this results results in a total of 1

输入：一张人像图片+待换衣服 输出：换上衣服的人像图片 整体模型：ACGPN 人体解析模型：Self Correction for Human Parsing 人体姿态模型：Alphapose 代码：https://github.com/Veronica1997/clothChange 欢迎大家star，关注，评论交流 ...