abcdvzz
abcdvzz
do i need to add a fc layer after the last conv layer when i write my own train.prototxt?
Could u pls give us a detailed configuration of the training of the CUB dataset on stage A? For example, input size, epoch, and transformation. I only get an accuracy...
So why do you change the lr_mult simultaneously?
No local-triplet loss means no vehicle part parser, no view-aware feature alignment, no common visible feature enhancement. So using a plain resnet50 with some tricks can get 77.2% on veri776....
So: training: train set, validation: test set, top-1 accuracy: test set. However, in the paper of veg-fru, top-1 accuracy is obtained in the val set. So there might be something...
> At this repository, we merely release the codes and models for testing. We will release the training codes soon. Dear Tianshui, when will u release the training codes? Thanks...
hello, can you give me a download URL of car-333 dataset? thank you very much.
> I solved this problem by regenerating the vocab.txt (use get_vocab.py), because this bug seems to occur when the drug can't tensorize by the vocab list. Can you elaborate more...
I got the same error, have you solved it?
> when you get 88.1% acc,did you train from finetune or just train from Scratch? finetune