iCub多传感器数据集

iCub多传感器数据集是由柏林洪堡大学计算机科学系自适应系统组创建的，用于机器人和计算机视觉应用的大型多传感器数据集。该数据集通过iCub机器人配备额外的深度传感器和彩色相机，采集了210个不同对象的颜色和深度信息，适用于对象识别、分类和动作识别等任务。数据集创建过程中，使用了多种传感器和自动化的数据采集管道，确保了数据的高质量和多样性。该数据集可用于多种认知机器人应用，如多模态对象概念形成和学习，以及在机器人和计算机视觉领域的应用，旨在解决多传感器信息融合的挑战。

The iCub Multi-Sensor Dataset is a large-scale multi-sensor dataset developed for robotics and computer vision applications, created by the Adaptive Systems Group at the Department of Computer Science, Humboldt-Universität zu Berlin. This dataset collects color and depth data of 210 distinct objects using an iCub robot equipped with additional depth sensors and color cameras, making it applicable to tasks including object recognition, classification and action recognition. During its development, multiple sensors and an automated data acquisition pipeline were employed to guarantee high data quality and diversity. This dataset supports a wide range of cognitive robotics applications, such as multimodal object concept formation and learning, as well as various robotics and computer vision use cases, with the goal of addressing the challenges associated with multi-sensor information fusion.