Online semi-supervised compressive coding for robust visual tracking

• Published in: Journal of visual communication and image representation 2014-07, Vol.25 (5), p.793-804
• Main Authors: Chen, Si, Li, Shaozi, Su, Songzhi, Cao, Donglin, Ji, Rongrong
• Format: Article
• Language: English
• Subjects: Algorithms; Applied sciences; Artificial intelligence; Coding; Compressive sensing; Computer science; control theory; systems; Data processing. List processing. Character string processing; Detection, estimation, filtering, equalization, prediction; Discriminative tracking; Exact sciences and technology; Generative tracking; Information, signal and communications theory; Mathematical models; Memory organisation. Data processing; Online; Online learning; Pattern recognition. Digital image processing. Computational geometry; Random projection; Semi-supervised coding; Semi-supervised compressive coding; Signal and communications theory; Signal, noise; Similarity; Software; State of the art; Telecommunications and information theory; Tracking; Visual; Visual tracking
• ISSN: 1047-3203; 1095-9076
• DOI: 10.1016/j.jvcir.2014.01.010

•We propose an online semi-supervised compressive coding method for visual tracking.•A novel adaptive compressive sensing based appearance model is presented.•An effective semi-supervised coding technique is developed for sample labeling.•A robust discriminative classifier is learned in the weighted compressed domain.•Our algorithm has superior tracking performance on challenging video sequences. In this paper we propose an online semi-supervised compressive coding algorithm, termed SCC, for robust visual tracking. The first contribution of this work is a novel adaptive compressive sensing based appearance model, which adopts the weighted random projection to exploit both local and discriminative information of the object. The second contribution is a semi-supervised coding technique for online sample labeling, which iteratively updates the distributions of positive and negative samples during tracking. Under such a circumstance, the pseudo-labels of unlabeled samples from the current frame are predicted according to the local smoothness regularizer and the similarity between the prior and the current model. To effectively track the object, a discriminative classifier is online updated by using the unlabeled samples with pseudo-labels in the weighted compressed domain. Experimental results demonstrate that our proposed algorithm outperforms the state-of-the-art tracking methods on challenging video sequences.