levirve / lsun-room Goto Github PK
View Code? Open in Web Editor NEW[ICPR 2018] Indoor Scene Layout Estimation from a Single Image.
Home Page: https://levirve.github.io/lsun-room
License: MIT License
[ICPR 2018] Indoor Scene Layout Estimation from a Single Image.
Home Page: https://levirve.github.io/lsun-room
License: MIT License
The results with the shared weights are not as good as the ones reported. Can you kindly share the original weights?
Could you provide the ground truth data for the raw image input?
thanks for your code ,when i run the lsun_room demo.py there is a AttributeError: module 'onegan' has no attribute 'utils',but i have installed the Onegan 0.5.1, and i have found a test_utils.py in the onegan test test_data file , could you suplly the utils tank you very much @leVirve @nafmichael
賣扣老師.py is missing
Can you guide me through, running this code in my local Jupyter notebook from start to end.
I find that many papers use Hedau dataset
Can you please upload the pretrained model.? I ll be thankful to you.
This is error I get when I try to run 0ython main.py .. OneGAN module does not have any repository utils that contain defination for to-var. Kindly help
thank for your code ,when i run your code as the readme said ,i just get the semantic segmentation map ,but idonot get the edge map or layout map,how to get it
File "/home/mj/PycharmProjects/lsun-room-master/datasets/lsunroom.py", line 49, in getitem
label = F.resize(label, self.target_size, Image.NEAREST)
File "/home/mj/anaconda3/envs/huangsz/lib/python3.8/site-packages/torchvision/transforms/functional.py", line 319, in resize
raise TypeError('img should be PIL Image. Got {}'.format(type(img)))
TypeError: img should be PIL Image. Got <class 'torch.Tensor'>
how to solve it thank you every much
Getting Error:
Can't find trainer.賣扣老師
Hi there,
I ran you code and had this problem.
Referring to other issues, I used OneGan 0.3.2 and v2018.jan version of this.
Any tips for help ?
Thank you for your code. Would you mind to share a model weight to try it please? I don't have a GPU powerful enought to train....
I tested the model on the Lsun dataset, but onlt get PE 19.3%. Is the problem of pretrained.ckpt or my wrong operation?
Can you please provide a link to the article which describes the algorithm? :)
Thanks for your great work!
As mentioned on README, the original LSUN Layout Challenge website is down, so It's really hard to collect the DB.
I found training/valid dataset from your project, but still it seeems there is no test DB anywhere.
Can you share the Test DB and also, if possible, can you share the LSUN original version of train/valid DB (.npz) ?
Thanks a million, in advance.
Thank You!
Originally posted by @Nethika in #3 (comment)
hello ,have you download the dataset,and how did you get the data
FileNotFoundError: [Errno 2] No such file or directory: '../data/lsun_room/training.mat
I have seen the OneGAN in code (no where mentioned in the paper) and its only used for metric calculation. could you please help me to understand how and why the OneGAN was used?
Or do you have a result like room layout image?
Could you share main.py usage example as it's getting difficult for me to understand arguments of it and also how layout_seg_images will be generated?
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.