两个问题
shoutOutYangJie opened this issue · comments
- 在测试阶段,使用的行人检测框架是?
- 从当前帧crop出的人体,扩大25%用于截取前一帧和后一帧的同一个人体,这是假设时序连贯。但你们如何能将整个视频的同一个人物的相同节点连贯起来? 这才是tracking的定义,论文中没有提到。
谢谢。
Thanks for your questions.
Problem 1
To ensure the fairness of the method comparison, we use the bounding box detection results provided by posewarper.
Problem 2
DCPose focuses on solving the human pose estimation for video and does not involve tracking.