GRIFFIN Perception Dataset
收藏资源简介:
GRIFFIN Perception Dataset是由西班牙塞维利亚大学的GRVC机器人实验室开发的数据集,专注于鸟类规模扑翼机器人的感知技术。该数据集包含21个子集,涵盖了三种不同的飞行场景(一个室内和两个室外)。数据集内容包括从事件相机、传统相机和两个惯性测量单元(IMUs)收集的数据,以及来自激光跟踪器或运动捕捉系统的地面实况测量。创建过程涉及在不同场景中收集多种类型的飞行数据。该数据集的应用领域主要集中在开发和评估基于图像和事件的感知技术,如特征提取、里程计和物体检测,旨在解决扑翼机器人在飞行中遇到的高振动水平和光照条件剧烈变化等问题。
The GRIFFIN Perception Dataset was developed by the GRVC Robotics Laboratory at the University of Seville, Spain, focusing on perception technologies for bird-scale flapping-wing robots. This dataset includes 21 subsets covering three distinct flight scenarios (one indoor and two outdoor). It contains data collected from event cameras, conventional cameras, two inertial measurement units (IMUs), as well as ground truth measurements from laser trackers or motion capture systems. The dataset creation process involved collecting diverse flight data across different scenarios. The primary application fields of this dataset center on developing and evaluating image-based and event-based perception technologies such as feature extraction, odometry and object detection, aiming to address challenges encountered by flapping-wing robots during flight, including high vibration levels and drastic changes in lighting conditions.
- 1The GRIFFIN Perception Dataset: Bridging the Gap Between Flapping-Wing Flight and Robotic PerceptionGRVC Robotics Lab Sevilla, Universidad de Sevilla, Spain · 2021年



