A Sparse Probabilistic Learning Algorithm for Real-Time Tracking

This paper addresses the problem of applying powerful pattern recognition algorithms based on kernels to efficient visual tracking. Recently Avidan [1] has shown that object recognizers using kernel-SVMs can be elegantly adapted to localization by means of spatial perturbation of the SVM, using optic flow. Whereas Avidan's SVM applies to each frame of a video independently of other frames, the benefits of temporal fusion of data are well known. This issue is addressed here by using a fully probabilistic `Relevance Vector Machine' (RVM) to generate observations with Gaussian distributions that can be fused over time. To improve performance further, rather than adapting a recognizer, we build a localizer directly using the regression form of the RVM. A classification SVM is used in tandem, for object verification, and this provides the capability of automatic initialization and recovery. The approach is demonstrated in real-time face and vehicle tracking systems. The `sparsity...

• Type: Conference
• Year: 2003
• Where: ICCV
• Authors: Oliver M. C. Williams, Andrew Blake, Roberto Cipolla
• Keywords: Classification Svm, Computer Vision, Efficient Visual Tracking, ICCV 2003, Powerful Pattern Recognition, Relevance Vector Machine, Temporal Fusion