mxnet_Realtime_Multi-Person_Pose_Estimation
mxnet_Realtime_Multi-Person_Pose_Estimation copied to clipboard
Details about CPM refinement mentioned in paper?
Table 4 in papar mentioned that CPM refinement can provide more than 2-point AP increase. Is it a post-processing method? Has it been added to this existed code?
The original paper has python demo and evaluation matlab code. They talked about the evaluation in original caffe implementation. The original evaluation code seems to be Matlab( I am trying to reimplement the evaluation code in Python). https://github.com/ZheC/Realtime_Multi-Person_Pose_Estimation/issues/72
If you are interested in it, we can collaborate in reimplementing the evaluation code in python. I am also interested to the performance of CPM refinement. @hufengshuo07
Hi @hufengshuo07, I finished the evaluation code. https://github.com/dragonfly90/mxnet_Realtime_Multi-Person_Pose_Estimation/blob/master/evaluation_coco.py
Have you finished the CPM refinement? @