22 February 2017 Self-paced model learning for robust visual tracking
Author Affiliations +
In visual tracking, learning a robust and efficient appearance model is a challenging task. Model learning determines both the strategy and the frequency of model updating, which contains many details that could affect the tracking results. Self-paced learning (SPL) has recently been attracting considerable interest in the fields of machine learning and computer vision. SPL is inspired by the learning principle underlying the cognitive process of humans, whose learning process is generally from easier samples to more complex aspects of a task. We propose a tracking method that integrates the learning paradigm of SPL into visual tracking, so reliable samples can be automatically selected for model learning. In contrast to many existing model learning strategies in visual tracking, we discover the missing link between sample selection and model learning, which are combined into a single objective function in our approach. Sample weights and model parameters can be learned by minimizing this single objective function. Additionally, to solve the real-valued learning weight of samples, an error-tolerant self-paced function that considers the characteristics of visual tracking is proposed. We demonstrate the robustness and efficiency of our tracker on a recent tracking benchmark data set with 50 video sequences.
© 2017 SPIE and IS&T
Wenhui Huang, Wenhui Huang, Jason J. Gu, Jason J. Gu, Xin Ma, Xin Ma, Yibin Li, Yibin Li, } "Self-paced model learning for robust visual tracking," Journal of Electronic Imaging 26(1), 013016 (22 February 2017). https://doi.org/10.1117/1.JEI.26.1.013016 . Submission: Received: 4 November 2016; Accepted: 26 January 2017
Received: 4 November 2016; Accepted: 26 January 2017; Published: 22 February 2017


Adaptive model MeanShift tracking
Proceedings of SPIE (March 12 2013)
Object tracking algorithm based on contextual visual saliency
Proceedings of SPIE (September 26 2016)

Back to Top