Research on unsupervised people re-identification based on k-means clustering
Huapeng Cai, Jianjun He, Li Zhe Luo, Yifan Wu · 2021
People re-identification is a direction in the field of computer vision. With the development of deep learning, pedestrian re-identification has become a popular research direction in recent years. At present, pedestrian re-identification has achieved good results in supervised learning with data annotation, but it still faces challenges in the unsupervised field. We propose a pedestrian re-identification method based on SE-ResNet50 k-means clustering and merging. The experimental results in the two large data setsMarket-1501 and Duke MTMC-reID have achieved better than the current mainstream Method for better results.