OWD

OWD数据集是由重庆大学微电子与通信工程学院开发的一个开放世界、多样化、跨时空的动态人物再识别基准。该数据集包含136,614条记录，覆盖多种场景，如街道、交叉口、购物中心等，以及从白天到夜晚的丰富光照变化。数据集通过多个摄像头在所有季节中部署，捕捉到多样的人物状态和外观，同时保护隐私，不显示面部。OWD数据集旨在解决现有数据集在多样性和动态场景适应性方面的不足，推动人物再识别模型在开放世界和动态野外的应用。

The OWD dataset is an open-world, diverse, cross-spatiotemporal dynamic person re-identification benchmark developed by the School of Microelectronics and Communication Engineering, Chongqing University. It contains 136,614 records, covering various scenarios including streets, intersections, shopping malls, etc., as well as rich illumination variations ranging from daytime to nighttime. Deployed across all seasons with multiple cameras, the dataset captures diverse person states and appearances while protecting privacy by excluding facial details. The OWD dataset is designed to address the limitations of existing datasets in terms of diversity and dynamic scene adaptability, and to promote the application of person re-identification models in open-world and dynamic wild environments.