VECTOR
收藏资源简介:
VECTOR数据集是由上海科技大学信息科学与技术学院移动感知实验室创建,旨在为多传感器同步定位与地图构建(SLAM)提供全面基准。该数据集包含多种传感器数据,包括事件立体相机、常规立体相机、多个深度传感器和惯性测量单元,所有数据均硬件同步并经过精确外部校准。数据集涵盖从小规模到大规模的环境,以及动态视觉传感器所针对的特定挑战。此外,数据集还提供了由高精度外部参考设备捕获的地面实况数据,如运动捕捉系统。VECTOR数据集适用于解决智能移动代理在复杂环境中的定位与地图构建问题,特别是在高动态或挑战性光照条件下。
The VECTOR Dataset was created by the Mobile Perception Laboratory, School of Information Science and Technology, ShanghaiTech University, aiming to provide a comprehensive benchmark for multi-sensor simultaneous localization and mapping (SLAM). This dataset contains various sensor data, including event stereo cameras, regular stereo cameras, multiple depth sensors, and inertial measurement units (IMUs). All data is hardware-synchronized and accurately externally calibrated. The dataset covers environments ranging from small-scale to large-scale, as well as specific challenges targeted by dynamic vision sensors. In addition, the dataset also provides ground truth data captured by high-precision external reference devices, such as motion capture systems. The VECTOR Dataset is suitable for addressing the localization and mapping problems of intelligent mobile agents in complex environments, especially under highly dynamic or challenging lighting conditions.




