Fast deformable structure regression tracking
Xian Yang, Shoujue Wang · IET Computer Vision · 2015
Visual object tracking is a challenging task because designing an effective and efficient appearance model is difficult. Current online tracking algorithms treat tracking as a classification task and use labelled samples to update appearance model. However, it is not clear to evaluate instance confidence belongs to the object. In this study, the authors propose a simple and efficient tracking algorithm with a deformable structure appearance. In their method, model updates with continuous labelled samples which are dense sampling. To improve the accuracy, they introduce a coupled‐layer regression model which prevents negative background from impacting on the model learning rather than traditional classification. The proposed deformable structure regression tracker runs in real time and performs favourably against state‐of‐the‐art trackers on various challenging sequences.