Di He

Master, Tianjin Normal University
Building 15, Tianjin Normal University, Xiqing District, Tianjin, China
Github | Blog | Zhihu | Weibo | Wechat

At present, an national invention patent has accepted as the first student inventor, and supported by the key project of graduate scientific research innovation project of Tianjin Normal University.

Research interest: computer vision and person re-identification.

news

Aug 9, 2022	Our paper Integration graph attention network and multi-centre constrained loss for cross-modality person re-identification is accepted by IET Computer Vision 2022!

selected publications

Integration graph attention network and multi-centre constrained loss for cross-modality person re-identification

Di He, Jingrui Zhang, Zhong Zhang, Shuang Liu, and Tariq S. Durrani

IET Computer Vision 2022

Abs HTML PDF

Cross-modality person re-identification is a challenging task due to the large visual appearance difference between RGB and infrared images. Existing studies mainly focus on learning local features and ignore the correlation between local features. In this paper, the Integration Graph Attention Network is proposed to learn the completed correlation between local features via the graph structure. To this end, the authors learn the coarse-fine attention weights to aggregate the local features by considering local detail and global information. Furthermore, the Multi-Centre Constrained Loss is proposed to optimise the feature similarity by constraining the centres of modality and identity. It simultaneously utilises three kinds of centre constraints, that is intra-identity centre constraint, modality centre constraint, and inter-identity centre constraint, in order to reduce the influence of modality information explicitly. The proposed method is evaluated on two standard benchmark datasets, that is SYSU-MM01 and RegDB, and the results demonstrate that the authors’ method achieves better performance than the state-of-the-art methods, for example, surpassing NFS by 4.8% and 6.0% mAP on the single-shot setting in All-search and Indoor-search modes, respectively.
CSPS

Memory Bank for Unsupervised Domain Adaptation Person Retrieval

Di He, Yu Lian, Peng Xia, Suoku Wei, and Zhong Zhang

In International Conference on Communications, Signal Processing and Systems Organizing Committee 2022

PDF
CSPS

Cloud Type Classification Using Multi-modal Information Based on Multi-task Learning

Yaxiu Zhang, jiazu Xie, Di He, Qing Dong, Jiafeng Zhang, Zhong Zhang, and 1 more author

In International Conference on Communications, Signal Processing and Systems Organizing Committee 2021

PDF