keypoint_rcnn_training_pytorch
keypoint_rcnn_training_pytorch copied to clipboard
Can the annotations include "keypoints" without "bboxes"?
Alex, thanks for your contributions!
I have trained a new model on my own dataset, but it doesn't show high accuracy on predicting the key points. I was considering if the prediction of boundary boxes has some impact on that of key points? So, can the annotations include only "keypoints" without "bboxes"?