lhc1224 / cross-view-ag Goto Github PK
View Code? Open in Web Editor NEWOfficial PyTorch Implementation of Learning Affordance Grounding from Exocentric Images, CVPR 2022
License: MIT License
Official PyTorch Implementation of Learning Affordance Grounding from Exocentric Images, CVPR 2022
License: MIT License
when I run test.py,a typeError came out,and I found that in model.py fucion MODEL is:
def MODEL(args, num_classes=36,
align_w=0.2, distil_w=0.1,
cls_w=1, pretrained=True,n=3,D=512):
but in test.py,
model = MODEL(args, num_classes=len(aff_list),
align_channels=256, align_w=0.5, distil_w=0.5,
cls_w=1, pretrained=False).cuda()
is the patameter align_channels irrelevant? shall we just delete it?
It had a result on pycharm IDE ,but not work on the server unless I delete that parameter.
By the way,can I ask for the code about plotting,I'm also a student from department of automation,ustc.
thank you for nice work(good task and dataset)π. Could you provide the model weights? please
Thanks for your great work.Before I try to train a model,I run process_gt.py to generate Unseen_gt.t7οΌand I replaced Unseen with Seen to generate Seen_gt.t7.
gt_root="dataset/AGD20K/Seen/testset/GT/"
torch.save(dict_1,"Seen_gt.t7")
but when I train a model,it didn't work,do you know what's the problem?()
Train begining!
train Epoch:[1/35] step:[50/1254] cls_loss: 7.113 dist_loss: 1.722 exo_acc: 22.75% ego_acc: 21.88%
train Epoch:[1/35] step:[100/1254] cls_loss: 4.066 dist_loss: 1.937 exo_acc: 30.00% ego_acc: 28.69%
train Epoch:[1/35] step:[150/1254] cls_loss: 5.073 dist_loss: 2.349 exo_acc: 35.25% ego_acc: 32.38%
train Epoch:[1/35] step:[200/1254] cls_loss: 3.748 dist_loss: 2.042 exo_acc: 37.56% ego_acc: 34.66%
train Epoch:[1/35] step:[250/1254] cls_loss: 2.537 dist_loss: 1.704 exo_acc: 39.25% ego_acc: 35.85%
train Epoch:[1/35] step:[300/1254] cls_loss: 4.081 dist_loss: 1.695 exo_acc: 40.94% ego_acc: 37.40%
train Epoch:[1/35] step:[350/1254] cls_loss: 3.147 dist_loss: 1.032 exo_acc: 42.61% ego_acc: 38.57%
train Epoch:[1/35] step:[400/1254] cls_loss: 2.145 dist_loss: 1.638 exo_acc: 43.98% ego_acc: 39.78%
train Epoch:[1/35] step:[450/1254] cls_loss: 5.803 dist_loss: 1.145 exo_acc: 45.15% ego_acc: 40.65%
train Epoch:[1/35] step:[500/1254] cls_loss: 3.255 dist_loss: 1.598 exo_acc: 46.45% ego_acc: 41.49%
train Epoch:[1/35] step:[550/1254] cls_loss: 1.971 dist_loss: 1.693 exo_acc: 47.32% ego_acc: 42.10%
train Epoch:[1/35] step:[600/1254] cls_loss: 1.749 dist_loss: 1.391 exo_acc: 48.07% ego_acc: 42.68%
train Epoch:[1/35] step:[650/1254] cls_loss: 2.975 dist_loss: 1.919 exo_acc: 48.81% ego_acc: 43.12%
train Epoch:[1/35] step:[700/1254] cls_loss: 2.258 dist_loss: 1.279 exo_acc: 49.47% ego_acc: 43.54%
train Epoch:[1/35] step:[750/1254] cls_loss: 2.977 dist_loss: 1.647 exo_acc: 49.87% ego_acc: 43.95%
train Epoch:[1/35] step:[800/1254] cls_loss: 2.014 dist_loss: 1.469 exo_acc: 50.30% ego_acc: 44.44%
train Epoch:[1/35] step:[850/1254] cls_loss: 2.583 dist_loss: 1.326 exo_acc: 50.64% ego_acc: 44.65%
train Epoch:[1/35] step:[900/1254] cls_loss: 2.099 dist_loss: 1.432 exo_acc: 51.14% ego_acc: 45.11%
train Epoch:[1/35] step:[950/1254] cls_loss: 2.247 dist_loss: 1.262 exo_acc: 51.71% ego_acc: 45.61%
train Epoch:[1/35] step:[1000/1254] cls_loss: 3.252 dist_loss: 1.017 exo_acc: 52.01% ego_acc: 45.85%
train Epoch:[1/35] step:[1050/1254] cls_loss: 1.915 dist_loss: 1.219 exo_acc: 52.30% ego_acc: 46.11%
train Epoch:[1/35] step:[1100/1254] cls_loss: 2.588 dist_loss: 1.692 exo_acc: 52.57% ego_acc: 46.27%
train Epoch:[1/35] step:[1150/1254] cls_loss: 2.674 dist_loss: 1.557 exo_acc: 52.88% ego_acc: 46.49%
train Epoch:[1/35] step:[1200/1254] cls_loss: 2.227 dist_loss: 1.067 exo_acc: 53.17% ego_acc: 46.79%
train Epoch:[1/35] step:[1250/1254] cls_loss: 2.756 dist_loss: 1.217 exo_acc: 53.41% ego_acc: 47.00%
Traceback (most recent call last):
File "train.py", line 221, in
masks = torch.load(args.divide+"_gt.t7")
File "/home/chenchangmao/anaconda3/envs/affordance/lib/python3.7/site-packages/torch/serialization.py", line 699, in load
with _open_file_like(f, 'rb') as opened_file:
File "/home/chenchangmao/anaconda3/envs/affordance/lib/python3.7/site-packages/torch/serialization.py", line 230, in _open_file_like
return _open_file(name_or_buffer, mode)
File "/home/chenchangmao/anaconda3/envs/affordance/lib/python3.7/site-packages/torch/serialization.py", line 211, in init
super(_open_file, self).init(open(name, mode))
FileNotFoundError: [Errno 2] No such file or directory: 'Seen_gt.t7'
hello, thank you for providing your work.I would like to ask, can you post the link to the testsetv2 dataset in the description?
@lhc1224
hello, thank you for providing your work.
i have confirmed your code with pretrained model(download the model through google drive)
by the way, size mismatch between shape from checkpoint and shape in current model occurs..
can you confirm your pretrained model with code?
A declarative, efficient, and flexible JavaScript library for building user interfaces.
π 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. πππ
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 β€οΈ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.