EgoReID

EgoReID数据集是由伊斯坦布尔技术大学和中佛罗里达大学合作创建，专门用于研究基于移动设备第一人称视角视频中的人物再识别。该数据集通过三部移动手机同步录制，覆盖非重叠视野，记录了约900个不同身份的10,200个轨迹，总计176,000次人物检测。数据集不仅包含视频数据，还附带了12种传感器元数据，如相机方向、旋转等，以支持更精确的分析。创建过程中，使用了先进的检测和跟踪技术，确保数据的准确性和丰富性。EgoReID数据集的应用领域主要集中在提升视频监控、活动分析和目标跟踪等领域的技术水平，特别是在处理由于相机移动、光照变化和背景复杂性带来的挑战方面。

The EgoReID dataset was collaboratively developed by Istanbul Technical University and the University of Central Florida, specifically for research on person re-identification in first-person perspective videos captured by mobile devices. This dataset is synchronously recorded using three mobile phones with non-overlapping fields of view, encompassing 10,200 trajectories across approximately 900 unique identities, with a total of 176,000 person detections. In addition to video data, the dataset also provides 12 types of sensor metadata such as camera orientation and rotation to enable more precise analytical studies. Advanced detection and tracking technologies were employed during the dataset creation process to ensure the accuracy and richness of the collected data. The primary application scenarios of the EgoReID dataset focus on advancing the technical levels of fields including video surveillance, activity analysis and target tracking, especially in tackling challenges caused by camera movement, illumination changes and complex background environments.