1 paper
Zhihan Kang, Boyu Wang
We propose SegVec3D, a novel framework for 3D point cloud instance segmentation that integrates attention mechanisms, embedding learning, and cross-modal alignment. The approach bu…