DCL

DCL copied to clipboard

在cuda11 torch1.8下运行会报ｒｕｎｔｉｍｅ:ｅｒｒｏｒ　forward static method错，在ｔｏｒｃｈ　１.２　ｒｔｘ　２０７０ｓ下运行是正常的，请给你指导和帮忙，谢谢

joe3000

Reproduce the results in the paper

1

In the paper, on CUB they mentioned 87.8% accuracy on CUB, but I can only get 86.9% using this repository.

mmaaz60

TypeError

1

我遇到了这样的错误，有大佬知道这是为啥吗，感谢感谢 TypeError: img should be PIL Image. Got

98Ray

sherlockers

When I train the model in 448 resolutions, it works fine. But when I set `--size 256 --crop 224`, Error occurs. How can I fix this error?

XiaoyuWant

about AIR dataset

1

i only achieved the 90.4% acc on AIR dataset, have anyone achieved 93%?

hyao1

Hi, I recently did some research on fine-grained classification, and I find that there are only train.txt and test.txt in your directory 'dataset'. Is the 'val.txt' in the code actually...

WindVChen

Specifically, where is the loc_loss in code? where is the adversarial loss? I can't find the regression between the predicted coordinates and original coordinates, where is that? Many thanks!

y6216886

assert Config.cls_2 ^ Config.cls_2xmul AssertionError

5

Namespace(auto_resume=False, backbone='resnet50', base_lr=0.0008, check_point=5000, cls_2=False, cls_lr_ratio=10.0, cls_mul=False, crop_resolution=448, dataset='CUB', decay_step=60, discribe='', epoch=360, resize_resolution=512, resume=None, save_point=5000, start_epoch=0, swap_num=[7, 7], train_batch=16, train_num_workers=16, val_batch=512, val_num_workers=32) **Traceback** (most recent call last): File "e:/graduation_project/object_detection-master/DCL-master/train.py", line...

rnlee1998

when I run the test.py, I got the error. How can I solve it ? ---- Namespace(acc_report=True, auto_resume=True, backbone='resnet50', batch_size=4, crop_resolution=448, dataset='SHIP', num_workers=0, resize_resolution=512, resume=None, save_suffix=None, swap_num=[7, 7], version='test') _42315_SHIP...

rnlee1998