Temporal Continuity Based Unsupervised Learning for Person Re-identification

Abstract

Most existing person re-identification (re-id) methods generally require a large amount of, difficult to collect, identity-labeled data to act as discriminative guideline for representation learning. To overcome this problem, we propose an unsupervised center-based clustering approach capable of progressively learning and exploiting the underlying re-id discriminative information from temporal continuity within a camera. We call our framework Temporal Continuity based Unsupervised Learning (TCUL). Specifically, TCUL simultaneously does center based clustering of unlabeled (target) dataset and fine-tunes a convolutional neural network (CNN) pre-trained on irrelevant labeled (source) dataset to enhance discriminative capability of the CNN for the target dataset. Furthermore, it exploits temporally continuous nature of images within-camera jointly with spatial similarity of feature maps across-cameras to generate reliable pseudo-labels for training a re-identification model. Extensive experiments on three large-scale person re-id benchmark datasets demonstrate superiority of TCUL over existing state-of-the-art unsupervised person re-id methods.