Multimodality Helps Few-Shot 3D Point Cloud Semantic Segmentation
Few-shot 3D point cloud segmentation (FS-PCS) aims at generalizing models to segment
novel categories with minimal annotated support samples. While existing FS-PCS methods …
novel categories with minimal annotated support samples. While existing FS-PCS methods …
Dynamic Taylor Convolutional Neural Network for Few-Shot Point Cloud Semantic Segmentation
C Wang, X Fang, LIU Zhonghang, Z Hu, X Gao… - openreview.net
Few-shot Point cloud semantic segmentation remains a challenge in the field of computer
vision due to the limitations of the pre-training learning paradigm and insufficient local …
vision due to the limitations of the pre-training learning paradigm and insufficient local …