Event-Based Crossing Dataset (EBCD)
收藏资源简介:
Event-Based Crossing Dataset (EBCD)是一个面向行人和车辆检测的动态户外环境综合数据集,由马里兰大学的研究团队创建。该数据集采用多阈值框架,通过在十个不同的阈值水平(4, 8, 12, 16, 20, 30, 40, 50, 60, 75)捕获事件基于图像,以评估对象检测在不同稀疏性和噪声抑制条件下的性能。数据集基于南安普顿大学行人数据集制作,包含约30000张图像,旨在推动事件基于视觉的低延迟、高保真度神经形态成像技术的发展。
Event-Based Crossing Dataset (EBCD) is a comprehensive dataset for pedestrian and vehicle detection in dynamic outdoor environments, created by a research team from the University of Maryland. This dataset adopts a multi-threshold framework, capturing event-based images at ten distinct threshold levels (4, 8, 12, 16, 20, 30, 40, 50, 60, 75) to evaluate object detection performance under varying sparsity and noise suppression conditions. Derived from the University of Southampton Pedestrian Dataset, it contains approximately 30,000 images and aims to advance the development of low-latency, high-fidelity neuromorphic imaging technologies based on event-based vision.




