SemanticKITTI 数据集，使用 LiDAR 序列进行语义场景理解的数据集

SemanticKITTI 是自动驾驶领域的权威数据集，它基于 KITTI 数据集，对 KITTI Vision Odometry Benchmark 中的所有序列都进行了标注，同时还为 LiDAR 360 度范围内采集到的所有目标，进行了密集的逐点注释。 该数据集包含 28 个标注类别，分为静态对象和动态对象，既包括行人、车辆等交通参与者，也包括停车场、人行道等地面设施。

SemanticKITTI is an authoritative dataset in the autonomous driving domain. Developed based on the KITTI dataset, it annotates all sequences within the KITTI Vision Odometry Benchmark, and additionally provides dense per-point annotations for every object captured within the 360° LiDAR scanning range. The dataset includes 28 annotated categories, divided into static objects and dynamic objects, covering both traffic participants such as pedestrians and vehicles, as well as ground facilities like parking lots and sidewalks.

搜集汇总

背景与挑战

背景概述

SemanticKITTI是自动驾驶领域的权威数据集，基于KITTI数据集，提供LiDAR 360度范围内所有目标的密集逐点注释，包含28个标注类别，涵盖静态和动态对象。