Beyond triplet loss: a deep quadruplet network for person re-identification

TY - GEN

T1 - Beyond triplet loss

T2 - IEEE International Conference on Computer Vision and Pattern Recognition

AU - Chen, Weihua

AU - Chen, Xiaotang

AU - Zhang, Jianguo

AU - Huang, Kaiqi

N1 - This work is funded by the National Key Research and Development Program of China (2016YFB1001005), the National Natural Science Foundation of China (Grant No. 61673375 and Grant No. 61403383), and the Projects of Chinese Academy of Science, (Grant No. QYZDB-SSWJSC006, Grant No.173211KYSB20160008).

PY - 2017

Y1 - 2017

N2 - Person re-identification (ReID) is an important task in wide area video surveillance which focuses on identifying people across different cameras. Recently, deep learning networks with a triplet loss become a common framework for person ReID. However, the triplet loss pays main attentions on obtaining correct orders on the training set. It still suffers from a weaker generalization capability from the training set to the testing set, thus resulting in inferior performance. In this paper, we design a quadruplet loss, which can lead to the model output with a larger inter-class variation and a smaller intra-class variation compared to the triplet loss. As a result, our model has a better generalization ability and can achieve a higher performance on the testing set. In particular, a quadruplet deep network using a margin-based online hard negative mining is proposed based on the quadruplet loss for the person ReID. In extensive experiments, the proposed network outperforms most of the state-of-the-art algorithms on representative datasets which clearly demonstrates the effectiveness of our proposed method.

AB - Person re-identification (ReID) is an important task in wide area video surveillance which focuses on identifying people across different cameras. Recently, deep learning networks with a triplet loss become a common framework for person ReID. However, the triplet loss pays main attentions on obtaining correct orders on the training set. It still suffers from a weaker generalization capability from the training set to the testing set, thus resulting in inferior performance. In this paper, we design a quadruplet loss, which can lead to the model output with a larger inter-class variation and a smaller intra-class variation compared to the triplet loss. As a result, our model has a better generalization ability and can achieve a higher performance on the testing set. In particular, a quadruplet deep network using a margin-based online hard negative mining is proposed based on the quadruplet loss for the person ReID. In extensive experiments, the proposed network outperforms most of the state-of-the-art algorithms on representative datasets which clearly demonstrates the effectiveness of our proposed method.

KW - Probes

KW - Training

KW - Testing

KW - Machine learning

KW - Feature extraction

KW - Loss measurement

U2 - 10.1109/CVPR.2017.145

DO - 10.1109/CVPR.2017.145

M3 - Conference contribution

SN - 9781538604588

SP - 1320

EP - 1329

BT - Proceedings of IEEE International Conference on Computer Vision and Pattern Recognition

PB - IEEE

Y2 - 21 July 2017 through 26 July 2017

ER -