Object tracking via collaborative multi-task learning and appearance model updating》.pdfVIP

Applied Soft Computing 31 (2015) 81–90 Contents lists available at ScienceDirect Applied Soft Computing j o u r na l h o me p ag e: w w w. elsev /locate/aso c Object tracking via collaborative multi-task learning and appearance model updating Xu Cheng, Nijun Li, Tongchi Zhou, Lin Zhou, Zhenyang Wu ∗ School of Information Science and Engineering, Southeast University, Nanjing 210096, PR China a r t i c l e i n f o a b s t r a c t Article history: In this paper, we propose a novel visual tracking algorithm using the collaboration of generative and Received 5 December 2013 discriminative trackers under the particle filter framework. Each particle denotes a single task, and we Received in revised form 8 February 2015 encode all the tasks simultaneously in a structured multi-task learning manner. Then, we implement Accepted 1 March 2015 generative and discriminative trackers, respectively. The discriminative tracker considers the overall Available online 7 March 2015 information of object to represent the object appearance; while the generative tracker takes the local information of object into account for handling partial occlusions. Therefore, two models are comple- Keywords: mentary during the tracking. Furthermore, we design an effective dictionary updating mechanism. The Visual tracking Particle filter dictionary is composed of fixed an