Person Re-identification in Crowded Scenes with Deep Learning

摘要

Person re-identification in crowded scenes is very important. Most images come from different surveillance video and cameras, and one person may look different in a variety of scenes, viewpoints, lighting and so on. The existing methods have limited effects in practical applications. In this paper, we propose a convolutional neural network for person re-identification in crowded scenes. The model structure of this network combines pedestrian detection and re-identification. In addition, we propose a loss function to better match the target person by calculating Pearson correlation evaluation. The experimental results show that our method is effective.