Binh-Son Hua

It was a server problem. I rebooted it. Thanks. 2021年4月17日(土) 20:37 坂本 ***@***.***>: > Hello, great work. When I download > http://103.24.77.34/2d3dmatch/3dmatch_0001.h5, the webpage shows Error: > EMFILE: too many...

Sorry for the trouble. Rebooted. I should have thought of a better way to host the files.. 2021年5月28日(金) 18:15 xiaoguang12 ***@***.***>: > I had received the same problem so i...

Just restarted the server. Could you try again? Thanks. 2021年11月24日(水) 15:31 Pterosaur-Yao ***@***.***>: > Thanks for your great work. I have faced with the same problem, so how can >...

I also have the same question: does SegNet actually use a mean value or mean file during its training? Does it make sense to define the DenseImageData layer (the top...

I got the same issue. It seems the newer versions of torchvision causes this problem. I switch to torchvision==0.2.1 as in the readme and the error is gone.

Please use this link instead: http://hkust-vgd.ust.hk/scenenn/ and http://hkust-vgd.ust.hk/scenenn/main/ I'll update the website soon. Thanks! 2022年6月2日(木) 9:31 bengio-ying ***@***.***>: > Hi, the backup link http://103.24.77.34/scenenn/ seems to be broken. I >...

Please use this backup link for the data: http://103.24.77.34:8080/scenenn/ Thanks. 2019年9月3日(火) 0:09 zzy : > The link > https://drive.google.com/drive/folders/0B-aa7y5Ox4eZWE8yMkRkNkU4Tk0 is > broken, as well as the download script. Could you...

There are two annotations available: the default is instance annotation, which is stored in the 'label' channel in the PLY file. 8844 represents the instance ID. You can look up...

Yes, the preprocessed data is from category annotation. It uses 76 scenes out of the 100 scenes from SceneNN, and follows the NYU-D 40 class annotation despite that some categories...

Thanks for finding this out. I should have the raw data of the 76 scenes. Let me find and get back to you in this thread by Wednesday night. 2019年9月3日(火)...