bharatsingh430 commented on issue #8156: Why mx-rcnn doesn't has proposal_target (cuda or C++) layer? URL: https://github.com/apache/incubator-mxnet/issues/8156#issuecomment-381365366 For end to end training, its not possible to prefetch. The position of the anchor-boxes changes based on the bounding box offsets in RPN and the assignment is done after the anchors are refined. The refined anchors (which are called proposals) are then compared with ground truth boxes, and label assignment is done (which is what proposal target layer does). I don't see how it can be done in pre-fetching
---------------------------------------------------------------- This is an automated message from the Apache Git Service. To respond to the message, please log on GitHub and use the URL above to go to the specific comment. For queries about this service, please contact Infrastructure at: us...@infra.apache.org With regards, Apache Git Services