IUST PersonReId
收藏资源简介:
IUST PersonReId数据集由伊朗科技大学计算机工程学院创建,旨在填补现有人员重识别数据集在伊斯兰文化背景下的空白。该数据集包含1847个身份,涵盖了伊朗和伊拉克的多个场景,如校园、市场、清真寺等,总计117,455张图像。数据集的创建过程包括从不同地点收集视频数据,并通过自动化跟踪算法进行标注,最终通过人工校正确保数据质量。该数据集的应用领域主要集中在提升在伊斯兰文化背景下的人员重识别模型的准确性和鲁棒性,解决因服装风格差异导致的模型性能下降问题。
IUST PersonReId Dataset was developed by the School of Computer Engineering, Iran University of Science and Technology, aiming to fill the gap in existing person re-identification datasets under the Islamic cultural context. This dataset contains 1847 identities, covering multiple scenarios in Iran and Iraq such as campuses, markets, mosques and other public places, with a total of 117,455 images. The dataset construction workflow includes collecting video data from various locations, annotating the data via automated tracking algorithms, and finally conducting manual verification to ensure data quality. The primary application fields of this dataset focus on enhancing the accuracy and robustness of person re-identification models under the Islamic cultural context, and addressing the problem of degraded model performance caused by differences in clothing styles.




