opencv pose keypoints explanation Needed

asked 2019-11-26 00:02:03 -0600

updated 2019-11-26 01:37:44 -0600

berak gravatar image

Hi, I have a doubt on opencv pose estimation module. I was using open pose boby_25 caffe prototext, here the final output key points representation is like 1*78*2*2 but when i m using this on python inference, i m getting like 1*76*46*46. can anyone explain this difference??

edit retag flag offensive close merge delete


I was using open pose boby_25 caffe prototext

do you have a link to that ? can we see your code ?

how / where do you get 1*78*2*2 ?

1*76*46*46 -- this would make sense for 25 part location (+ 1 bg) and 2x25 (xy) part affine maps

berak gravatar imageberak ( 2019-11-26 01:38:46 -0600 )edit