wis3D error

[dadawdadD]

Thank you for your excellent work! However, I have encountered some issues while reproducing the data generation pipeline.


Is there a problem with the SpatialRGPT/dataset_pipeline I reproduced? The frame in the image does not have a corresponding point cloud.

Additionally, the 3D bounding boxes in the images also seem to be incorrect, but I followed your steps and don’t know where the mistake lies.
Finally, I would like to ask if the xyxy in the json file generated in demo_out does not correspond to the target box in the image.
thanks!