Dual semantic interdependencies attention network for person re‐identification. Issue 25 (12th November 2020)

• Record Type: Journal Article
• Authors: Yang, Shengrong; Hu, Haifeng; Chen, Dihu; Su, Tao
• Is Part Of: Electronics letters. Volume 56:Issue 25(2020)
• Page Start: 1411
• Page End: 1413
• Publication Date: 2020-11-12
• Subjects: mathematics computing -- linear algebra -- recurrent neural nets -- convergence of numerical methods; linear algebraic systems -- recurrent neural network -- finite‐time convergence -- exponential term -- Zhang neural network dynamical system -- error‐monitoring function -- varying‐parameter finite‐time zeroing neural network -- convergence; Electronics -- Periodicals; 621.381
• DOI: 10.1049/el.2020.1786
• Languages: English
• ISSNs: 0013-5194

Abstract

Abstract : Attention mechanisms are widely used in re‐identification (reID) tasks, but few attention‐based architectures have considered integrating local features with their global dependencies, that is the previous works do not model the semantic interdependencies in both spatial dimension and channel dimension. Intuitively, for person feature representations, it is important to model the interdependencies between human body semantics. In this Letter, the authors proposed a dual semantic interdependencies attention module to capture semantic interdependencies in both spatial dimension and channel dimension simultaneously. Besides, they designed a deep supervision branch to directly guide the training of the attention modules and innovatively introduce a channels random dropping mechanism in the training phase to promote the attention modules to capture more discriminative information. Extensive experimental results show that the network merging the above strategies achieves state‐of‐the‐art results on the mainstream reID data sets.