3d-crowdpose-estimation-based-on-mvg
3d-crowdpose-estimation-based-on-mvg copied to clipboard
Maybe something wrong with the 2D annotations?
Thanks for the great work and for releasing the annotations!
But I found there exist some wrong annotations in Wildtrack dataset after I visualize it.
For example, in C1 00000050.png
, there are some redundant detections.
In C1/00000380.png
, the feet keypoints were labeled incorrectly.
Other frames have similar wrong annotations.
你好,这个数据的链接失效了,请问可以分享一下吗,谢谢
Hi dude, can you share annotations with us?