jayqine / amr Goto Github PK
View Code? Open in Web Editor NEWLicense: Apache License 2.0
License: Apache License 2.0
FileNotFoundError: [Errno 2] No such file or directory: 'sess/res50_cam.pth.pth'
FileNotFoundError: [Errno 2] No such file or directory: 'sess/res50_irn.pth.pth'
I also want to know the complete training steps, Thanks.
Hi author,
I run your code for about 10 days on 3090. and the result of eval_cam is only 0.428.
Is there anything wrong with my running command? I was running with "python run_sample.py --train_cam_pass True --train_amr_pass True --make_cam_pass True --eval_cam_pass True"
Cheers,
yuhao
Can you download the code from Github project and check out?
Maybe, this paper propose no new methods for WSSS. Your improvement is brought by cvpr2020 anti-adv method.
Hi, I have problem about that:
UserWarning: Using a non-full backward hook when the forward contains multiple autograd Nodes is deprecated and will be removed in future versions. This hook will be missing some grad_input. Please use register_full_backward_hook to get the documented behavior.
please help!
Could you explain why the code use flip operation in the lines 67 through 69? Several papers hava also used this operation.
cam2 = F.conv2d(x2, self.classifier2.weight)# 1*1 conv cam2 = F.relu(cam2) cam2 = cam2[0] + cam2[1].flip(-1)
Hello,
thanks for the great job!
I am confusing about the model provided. Is it generated from segmentation stage or classification phase?
Thank you!
hello,i wonder how get cam image after the step 'train cam'
Hello! Thanks for your great work!
I am confused about some lines in net/resnet50_amr.py. Did you mean that the spotlight branch is not supervised by classification? Is this equal to with torch.no_grad()
?
Lines 47 to 58 in de39168
Looking forward to your reply.
Hello, there was an exception occurred when I run eval script with command line python run_sample.py --make_cam_pass True --eval_cam_pass True
.
TypeError: forward() got an unexpected keyword argument 'step'
The forward
method of the class CAM
seemingly did not support a parameter named step
.
But the make_came
step had provided it.
I'm wondering what's the effect of step
, and what should I do for the code.
the cam making stage is too slow
I have encountered such an error:
Traceback (most recent call last):
File "run_sample.py", line 114, in
step.train_amr.run(args)
File "/home/stuc/zhx/AMR-main/step/train_amr.py", line 49, in run
train_dataset = voc12.dataloader.VOC12ClassificationDataset(args.train_list, voc12_root=args.voc12_root,
File "/home/stuc/zhx/AMR-main/voc12/dataloader.py", line 167, in init
super().init(img_name_list_path, voc12_root,
File "/home/stuc/zhx/AMR-main/voc12/dataloader.py", line 115, in init
self.img_name_list = load_img_name_list(img_name_list_path)
File "/home/stuc/zhx/AMR-main/voc12/dataloader.py", line 60, in load_img_name_list
img_name_list = np.loadtxt(dataset_path, dtype=np.int32)
File "/home/stuc/.conda/envs/amr/lib/python3.8/site-packages/numpy/lib/npyio.py", line 1308, in loadtxt
arr = _read(fname, dtype=dtype, comment=comment, delimiter=delimiter,
File "/home/stuc/.conda/envs/amr/lib/python3.8/site-packages/numpy/lib/npyio.py", line 979, in _read
arr = _load_from_filelike(
ValueError: could not convert string '2007_000032' to int32 at row 0, column 1.
Looking forward to your reply,thank you!!!!
Hello, I want to ask which segmentation code did you use to train the pseudo-label?
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.