Abstract
Person reidentification is aimed at solving the problem of matching and identifying people under the scene of cross cameras. However, due to the complicated changes of different surveillance scenes, the error rate of person reidentification exists greatly. In order to solve this problem and improve the accuracy of person reidentification, a new method is proposed, which is integrated by attention mechanism, hard sample acceleration, and similarity optimization. First, the bilinear channel fusion attention mechanism is introduced to improve the bottleneck of ResNet50 and fine-grained information in the way of multireceptive field feature channel fusion is fully learnt, which enhances the robustness of pedestrian features. Meanwhile, a hard sample selection mechanism is designed on the basis of the P2G optimization model, which can simplify and accelerate picking out hard samples. The hard samples are used as the objects of similarity optimization to realize the compression of the model and the enhancement of the generalization ability. Finally, a local and global feature similarity fusion module is designed, in which the weights of each part are learned through the training process, and the importance of key parts is automatically perceived. Experimental results on Market-1501 and CUHK03 datasets show that, compared with existing methods, the algorithm in this paper can effectively improve the accuracy of person reidentification.
Highlights
As an important intelligent video analysis technology, person reidentification is widely used in the fields of intelligent security, case detection, lost query, intelligent interaction, and so on
Based on the P2G similarity optimization model, the hard sample mining and feature grouping and similarities fusion module are introduced to improve the accuracy of person reidentification model. e whole model can be divided into three modules: feature extraction module, hard sample mining module, and feature group and similarities optimization fusion module
In view of the large difference of person under different visual fields, in order to fully extract effective person details information and to learn robust person expression and solve the key problem of reidentification task, this paper proposes a person reidentification algorithm, which includes three main aspects
Summary
As an important intelligent video analysis technology, person reidentification is widely used in the fields of intelligent security, case detection, lost query, intelligent interaction, and so on. Aiming at extracting more robust feature for person images, based on the Siamese Network, we improve the convolution settings on the ResNet, integrate multiscale visual field features, and design the attention mechanism of channel fusion to enhance the expression of fine-grained feature information in the image. Unlike the existing feature measurement and rerank algorithms, based on the P2G similarity optimization model [21], a hard sample selection mechanism is set up, which uses the hard samples to optimize the P2G similarity to enhance the learning and generalization ability of the model. Experiments on Market-1501 and CUHK03 datasets show that the proposed model can extract person features completely and achieve higher recognition accuracy
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
