Unsupervised Person Re-identification by Deep Learning Tracklet Association
This addresses the need for scalable person re-identification in surveillance without manual labeling, though it is incremental as it builds on existing unsupervised methods.
The paper tackles the scalability problem in person re-identification by proposing an unsupervised deep learning approach that incrementally discovers discriminative information from automatically generated tracklet data, achieving state-of-the-art results on six benchmarking datasets.
Mostexistingpersonre-identification(re-id)methods relyon supervised model learning on per-camera-pair manually labelled pairwise training data. This leads to poor scalability in practical re-id deployment due to the lack of exhaustive identity labelling of image positive and negative pairs for every camera pair. In this work, we address this problem by proposing an unsupervised re-id deep learning approach capable of incrementally discovering and exploiting the underlying re-id discriminative information from automatically generated person tracklet data from videos in an end-to-end model optimisation. We formulate a Tracklet Association Unsupervised Deep Learning (TAUDL) framework characterised by jointly learning per-camera (within-camera) tracklet association (labelling) and cross-camera tracklet correlation by maximising the discovery of most likely tracklet relationships across camera views. Extensive experiments demonstrate the superiority of the proposed TAUDL model over the state-of-the-art unsupervised and domain adaptation re- id methods using six person re-id benchmarking datasets.