ma-xu/pointMLP-pytorch

part segmentation

Closed this issue · 5 comments

junshengzhou commented

Hi,
Thank you for sharing your code! But I have some difficult in training ShapeNetPart segmentation. I found that there is no folder named 'util' under 'part_segmentation' , so I can't find 'to_categorical', 'PartNormalDataset' etc. And I guess Class 'ShapeNetPart' might be another version of 'PartNormalDataset', but it seems to have different parameter.

ma-xu commented

@junshengzhou Thanks for your remind. Will double check and fix it soon.

swzaaaaaaa commented

There is indeed this problem. Has it been solved?

ma-xu commented

@swzaaaaaaa See #9 (comment)

We will update it after Mar/10 because we are busy with another project recently.

ma-xu commented

@swzaaaaaaa @junshengzhou
Thanks for your interest. I just double-checked: mistakenly uploaded the wrong folder in experiments.
Our development codes are based on PAConv, GDANet(yes), and Curevenet. We tried several implementations; as a result, we uploaded the wrong folder. You can see it in my pointsmlp repo, which is our development repo.
It is fixed now. Please have a try and let me know if there are still any issues.

ma-xu commented

@swzaaaaaaa @junshengzhou I will close this issue since there is no further discussion. Feel free to reopen it if necessary.