How to inference on my own point cloud data?

[jackchinor]

Nice works , I want to inference on my own point cloud data? what data should I prepare? Thanks.

[filaPro]

You need point cloud data in [[x, y, z, r, g, b], ...] format and superpoint clusterinzation of it. Unfortunately this second part is not straightforward and can be done with scannet segmentator or superpoint_transformer.

[jackchinor]

You need point cloud data in [[x, y, z, r, g, b], ...] format and superpoint clusterinzation of it. Unfortunately this second part is not straightforward and can be done with scannet segmentator or superpoint_transformer.

thanks. Need any other data? such as .pkl file. Could you give an script example? Thanks very much

[Nacriema]

Hi, currently I can create my own ply files that can work with the ScanNet's Segmentator.

[jackchinor]

@Nacriema Can you inference on your own ply data?

[Nacriema]

Yes, I can run my own ply data with ScanNet's Segmentator. But with the Superpoint transformer I still need more effort to investigate with the 3D Coordinates because I meet the error:

RuntimeError: Trying to create tensor with negative dimension -29: [1369933, -29]

when trying to perform segmentation with the original kitti-360 dataset.

[jackchinor]

@Nacriema I mean inference your data through unidet3d and output 3d object detection result

[Nacriema]

At the moment, the status is no. I am on the way to perform superpoint transformation like the author's suggestion.

[jackchinor]

@filaPro Need any other data? such as .pkl file. Could you give an script example or tips? Thanks very much

[filaPro]

It is possible to initialize your custom dataset class even without .pkl files. They contain mainly just list of paths to point clouds and annotations. But we also provide scripts, how to make them for our datasets in data directory.

[jackchinor]

It is possible to initialize your custom dataset class even without .pkl files. They contain mainly just list of paths to point clouds and annotations. But we also provide scripts, how to make them for our datasets in data directory.

Thanks, but I still don't know how to make a inference on my own point cloud data and super point data, seems it doesn't work when I run tools/test.py . Need your help, thanks