TrafficMOT
收藏资源简介:
TrafficMOT是一个专为复杂交通场景中多目标跟踪设计的新型基准数据集,由剑桥大学创建。该数据集包含2102个视频,覆盖印度多个城市的交通情况,具有高密度和多样性的交通场景。数据集通过固定CCTV摄像头收集,包含多种天气条件和光照情况,旨在解决现有数据集在模拟真实交通场景复杂性方面的不足。TrafficMOT不仅提供大量对象类别,还捕捉了交通场景中的复杂动态和拥堵情况,适用于推动交通监控和多目标跟踪技术的发展。
TrafficMOT is a novel benchmark dataset tailored for multi-object tracking in complex traffic scenarios, developed by the University of Cambridge. It comprises 2102 video clips covering traffic conditions across multiple cities in India, featuring high-density and highly diverse traffic scenes. Collected via fixed CCTV cameras, the dataset includes diverse weather conditions and lighting scenarios, aiming to address the shortcomings of existing datasets in simulating the complexity of real-world traffic scenarios. In addition to providing a wide range of object categories, TrafficMOT also captures the intricate dynamics and congestion within traffic scenarios, making it suitable for advancing the development of traffic monitoring and multi-object tracking technologies.

- 1TrafficMOT: A Challenging Dataset for Multi-Object Tracking in Complex Traffic Scenarios剑桥大学 · 2023年



