-
微博内容精选
-
Hi, I run the code as: python train.py --config_file configs/VeRi/vit_transreid_stride.yml MODEL.DEVICE_ID "('0')" and don't change the arguments except batch_size, because of limited GPU. I se…
-
请问车辆再识别的程序还会放上来嘛
-
Hello!
I have some trouble with following the following code - python generate_pkl.py veri776 --input-path --output-path ../outputs/veri776.pkl
What shall I put as the as the veri-776 dataset is …
-
我是一名刚读研的同学，目前研究方向是车辆重识别，想要复现一下这篇文章中的模型，但是自己知识还不太够，所以创建了一个QQ讨论组，想和大家一起完成文章的复现。群号是：1136870868。欢迎各位进群讨论
-
Hi,
I found that [euclidean distance](https://github.com/Xiangyu-CAS/AICity2020-VOC-ReID/blob/master/lib/utils/post_process.py#L37) was used in your code to calculate distmat, while it was cosine sim…
-
=> Start training
log_probs.shape = torch.Size([32, 4556])
targets.shape = torch.Size([32])
Traceback (most recent call last):
File "scripts/train.py", line 52, in
test_only=False
Fil…
-
你好，我用VeRi-776数据集进行训练，其中576类作为训练集，200类作为测试集。
1、在model.py中，将class_num=576。
2、在test.py中，将这里model_structure = ft_net(200)，设为测试集的200类，执行python test.py时会报错，说训练模型为576维的，而这里200维，不相符。 但是，将这里model_structure…
-
Hi, could you please clarify whether the results on VeRi-776 in your paper are obtained using image-to-track or image-to-image test fashion?
-
FloatingPointError: Loss became infinite or NaN at iteration=1!
loss_dict = {'loss_cls': tensor(nan, device='cuda:0', grad_fn=), 'loss_triplet': tensor(nan, device='cuda:0', grad_fn=)}
这是什么原因阿