Abstract
Nowadays, visual object tracking becomes a hotspot and difficulty to achieve a real-time and accurate target tracking, but the Siamese network has solved these difficulties because of its good tracking effect and real-time performance. The location of the target in the previous frame is the template, and the similarity matching is carried out in the search area of the current frame. However, it uses Alexnet network with simple structure and fewer layers to extract features, and just uses a score map to predict the final position of the object. Aiming at these problems, in this paper, we propose the Siamese network of fused response map that use the Alexnet network with fine tuning to extract target features, and weight fusion of score maps to estimate the final position of object. Sufficient experiments on the VOT2015 and OTB100 benchmarks validate that our tracker can improve tracking performance, and perform at 60FPS.
Highlights
Image tracking is to locate the object captured by the camera in some ways
In this paper, aiming at the shortcomings in traditional algorithms SiamFC, this paper proposes an object tracking algorithm based on siamese network of score map fusion
Experiments on two datasets: OTB100 and VOT2015, demonstrate improved performance compared with SiamFC
Summary
Image tracking is to locate the object captured by the camera in some ways. So, the object tracking firstly need to initialize the first frame of the entire video, mark the target to be tracked in the first frame, and predict the location and size of the object in the frame. Bo Li et al refer to the idea of SiamFC, combine RPN(Region Proposal Network) consisting of classification module and regression module with FC(Fully Convolution Network) network, and propose SiamRPN [29], which achieves effective tracking, and greatly improves the position detection accuracy of the target because of the regression ability of RPN. The SAMF introduce a multiple scales searching strategy by using a scaling pool approach to achieve adaptive tracking of target Both of the above methods use the scale pool method to solve the problem of target size change. The response map corresponding to the maximum value is used to estimate the target position, but for the tracking of any target, some target sizes change sharply, so in the Large-scale Single Object Tracking (LaSOT) [33], the accuracy and success rate of the above two algorithms are greatly reduced.
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.
