Preprocessed file are not generated in sdf_hand and sdf_obj but only in norm

[karta2155802]

Preprocessed file are not generated in sdf_hand and sdf_obj but only in norm

[zerchen]

Hi,

Previously, I also struggled with this issue, which probably suggests your PC lacks enough OpenGL support. If you work on a remote server, you need to execute "export PANGOLIN_WINDOW_URI=headless://". Sometimes, it still does not work on a remote server. I suggest to work on a local ubuntu 20 machine with a GPU and enough OpenGL support.

[WeedsInML]

When I generate sdf_hand and sdf_obj, it's so time-consuming. Could you please provide your data?

[zerchen]

The processed training files are large (about 460G). I will try to find ways to upload them into the cloud.

[CODE-SUBMIT]

Hi,

Previously, I also struggled with this issue, which probably suggests your PC lacks enough OpenGL support. If you work on a remote server, you need to execute "export PANGOLIN_WINDOW_URI=headless://". Sometimes, it still does not work on a remote server. I suggest to work on a local ubuntu 20 machine with a GPU and enough OpenGL support.

I only get norm folder, the error is terminate called after throwing an instance of 'pangolin::WindowExceptionNoKnownHandler, I check my ogengl, but there is nothing wrong.

How to solve it.

[zhangxinkang]

Hi,

Previously, I also struggled with this issue, which probably suggests your PC lacks enough OpenGL support. If you work on a remote server, you need to execute "export PANGOLIN_WINDOW_URI=headless://". Sometimes, it still does not work on a remote server. I suggest to work on a local ubuntu 20 machine with a GPU and enough OpenGL support.

Hi, i get files in norm, sdf_hand and sdf_obj folders, but the number of files in them are different.
In my local computer, i got about 50000 files in norm folder, but just about 30000 files in sdf_hand and sdf_obj.
In our ubuntu server, the number of files in sdf_hand is 1/10 of the number of files in norm.

Do you know how to solve this?

[zerchen]

Sorry, I have no idea of this issue. However, I would put my processed data on the google cloud around end of November.

[zerchen]

Hi,

I have put my preprocessed data online. You could have a check at https://drive.google.com/drive/folders/1GoaA6vB6TwAAHmaobVo5GjRoCq2wT21R. Hope it helps resolve your issues.

Best,
Zerui

[EAST-J]

Hi,

I have put my preprocessed data online. You could have a check at https://drive.google.com/drive/folders/1GoaA6vB6TwAAHmaobVo5GjRoCq2wT21R. Hope it helps resolve your issues.

Best, Zerui

Hi, Thanks for sharing. Could you also put preprocessed data for DexYCB?

[zerchen]

Yes, I will put the processed data for DexYCB next week.