There's something wrong with this about sucking #1
Description
Thank you for the great work！
Your documentation is so detailed that I, a beginner, can get up and running very quickly.
I didn't use the test data you provided; instead, I used point clouds from my own scenario as input. Also, I didn't set up the environment needed for generating data; I merely used the inference interface to generate suction location information on the point cloud. I have a few questions:
1.Does predicting suction locations only support input from a single object's point cloud, and not from the entire scene's point cloud?
2.Will the model be affected by the units of the point cloud?
3.Can the model understand the point cloud scene and provide suction location information for the best object to suction in the scene?