ASAP-Net: Attention and Structure Aware Point Cloud Sequence Segmentation

Recent works of point clouds show that mulit-frame spatio-temporal modeling outperforms single-fra