Point cloud completion using multiscale feature fusion and cross-regional attention
作者：
Highlights：
• An adaptive neighborhood query algorithm that acquires balanced local regions for partial point cloud.
• A cross-regional attention module that learns relationships among local regions under shared global conditions.
• A two-step decoder that generates complete point clouds based on cross-region features from coarse to fine.
• Better performance on simulated and real-world data, a comprehensive relation between local areas of inputs and outputs.
摘要
•An adaptive neighborhood query algorithm that acquires balanced local regions for partial point cloud.•A cross-regional attention module that learns relationships among local regions under shared global conditions.•A two-step decoder that generates complete point clouds based on cross-region features from coarse to fine.•Better performance on simulated and real-world data, a comprehensive relation between local areas of inputs and outputs.
论文关键词：3D vision,Point cloud completion,Conditional local feature,Cross-regional attention
论文评审过程：Received 18 December 2020, Revised 22 April 2021, Accepted 27 April 2021, Available online 30 April 2021, Version of Record 26 May 2021.
论文官网地址：https://doi.org/10.1016/j.imavis.2021.104193
