Chen Ma
Chen Ma
@arya-none @asa008 Do you solve this problem? I encounter the same issue!
I posted an issue here: https://github.com/microsoft/MMdnn/issues/925
It seems that this is a bug that caused by Fused Batch Normalization layer, but I did not use that layer. I only use `tf.layers.batch_normalization` as the BN layer in...
I think the memory leak due to RoI pooling layer, because when I copy the code of RoI pooling layer to my another project. It also memory leak on GPU.
@xdever I think if you calculate all pairs of boxes IOU first, then just for-loop once will ultimately boost speed, there have some trick in it, just see the source...
@affromero I am also working on this area. I think the paper he report is not correct. I am using end-to-end LSTM which cannot achieve 66% F1-score, that is too...
@affromero Which file is the train file and are you runing?
@affromero @YXWu-bjtu I think the F1 score reported in that paper has problem. My new research will be coming soon, it achieved F1 score of 63.1% on BP4D 3-fold test....
@affromero @YXWu-bjtu Although the author present a great paper, but unfortunately, the result of LSTM is incorrect of PAPER, it will not surpass 63% in BP4D of F1 score. I...
The following file is the compressed zip file of pytorch version of AU R-CNN. [AU_RCNN_pytorch.zip](https://github.com/machanic/AU_R-CNN/files/13696108/AU_RCNN_pytorch.zip)