Open ethanfetaya opened 8 years ago
Hi, we cannot release that code for the time being because our internal implementation is different from the public caffe and we do not use python. We observered joint training acheives accuracy on par with 4-step training of R-FCN, when OHEM is not performed.
@ethanfetaya what about your mAP?
Dear Dai, I tried to train the network end-to-end (i.e. add region proposal part similarly to faster-rcnn) but the results are not nearly as good as the ones presented in your paper (although not junk so it isn't an obvious bug). I assume my implementation isn't correct. Can you please release the full training code you used (i.e. region proposal and detection)?
Thanks, Ethan