Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

in_channels meaning in cfg #168

Open
haonan-duan opened this issue Jan 20, 2025 · 1 comment
Open

in_channels meaning in cfg #168

haonan-duan opened this issue Jan 20, 2025 · 1 comment

Comments

@haonan-duan
Copy link

Hello, thanks for your amazing project!

I want to use pretrained PointNeXt-XL model to extract point cloud features in my task. I am wondering what is the meaning of in_channels in cfg? I just found in_channels can be modified to any number even I have loaded the pretrained model, and its value is set to be 4 in default in every cfg file. Shouldn't input point clouds have 3 channels? And is there any influenece if I change in_channels from 4 to be 3? Thanks for your help!

@HuayuanSong
Copy link

HuayuanSong commented Feb 10, 2025

To my knowledge, the 4th input channel stems from the data augmentation, specifically, the height appending operation -'heights' that you can see being passed under the 'feature_keys' param in https://github.com/guochengqian/PointNeXt/blob/master/cfgs/shapenetpart/default.yaml .

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants