LiDAR video object segmentation with dynamic kernel refinement

Open Webpage

Jianbiao Mei, Yu Yang, Mengmeng Wang, Zizhang Li, Jongwon Ra, Yong Liu

Published: 01 Jan 2024, Last Modified: 07 May 2025Pattern Recognit. Lett. 2024EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Highlights•LiDAR-based VOS task, which segments specific 3D targets in LiDAR sequences.•KITTI-VOS dataset constructed from SemanticKITTI.•Memory- and tracking-based baselines for a comprehensive study.•Effective refine module to boost these baselines’ performances.