CID
CID copied to clipboard
kennethwdk
Code for "Contextual Instance Decoupling for Robust Multi-Person Pose Estimation", CVPR 2022 Oral
Hello, thanks for your excellent research. I have reproduced CID on the COCO dataset. However, the reproduced results on COCO val dataset are only 66.7% AP (Python 3.8.0, CUDA 11.4)....
Could you please tell me how to draw the picture like Figure 4 in your paper? Are you using any packages or python code? I would appreciate it if you...
Whether the val batch can only be set to 1 ? I would appreciate it if the author could answer these two questions!!!
Hello, your work is very worthy of our study. I have read your paper on Contextual Instance Decoupling for Instance-Level Human Analysis. This is an article worth learning, I would...
May I ask why all the detection results are zero after I switched the backbone and trained on COCO? Could you please share the training logs from your training session?
Hello, your work is very worthy of our study. I would like to ask whether the heatmaps of keypoints is obtained through annotation files or generated by keypoint data itself?...
I don't know why when I train the model on Ochuman or Mscoco, some errors in dataset loading occurs as below, [ WARN:[email protected]] global /io/opencv/modules/imgcodecs/src/loadsave.cpp (239) findDecoder imread_('data/ochuman/images/005034.jpg'): can't open/read...
Hi, thanks for your excellent work. The links of trained models seem unavailabel. Can you provide alternative download links?
Metadata
Owner
Metadata
Code for "Contextual Instance Decoupling for Robust Multi-Person Pose Estimation", CVPR 2022 Oral