1 paper
Xingyu Miao, Haoran Duan, Yang Bai +5
In this work, we propose a method that leverages CLIP feature distillation, achieving efficient 3D segmentation through language guidance. Unlike previous methods that rely on mult…