“…The aim of a visual tracking system is to locate a predefined target object on every frame of a video sequence. Automatic systems span a wide range of applications, such as traffic monitoring (Kamijo et al 2000;Hsieh et al 2006), surveillance (Haritaoglu et al 2000;Collins et al 2001), video retrieval and summarization (Luo et al 2003), vehicle navigation (Hashima et al 1997;Fraundorfer et al 2007), driver assistance (Handmann et al 1998;Avidan 2004), human computer interaction (Wren et al 1997;Liwicki and Everingham 2009) and face analysis (Gunes and Pantic 2010;Cohn et al 1999). Many tracking algorithms indicate that an adaptive approach based on online learning is advantageous to fixed appearance models learned offline (Babenko et al 2011;Ross et al 2008;Mei and Ling 2009).…”
Principal Component Analysis (PCA) is perhaps the most prominent learning tool for dimensionality reduction in pattern recognition and computer vision. However, the 2 -norm employed by standard PCA is not robust to outliers. In this paper, we propose a kernel PCA method for fast and robust PCA, which we call Euler-PCA (e-PCA). In particular, our algorithm utilizes a robust dissimilarity measure based on the Euler representation of complex numbers. We show that Euler-PCA retains PCA's desirable properties while suppressing outliers. Moreover, we formulate Euler-PCA in an incremental learning framework which allows for efficient computation. In our experiments we apply Euler-PCA to three different computer vision applications for which our method performs comparably with other stateof-the-art approaches.
Electronic supplementary material The online version of this article
“…The aim of a visual tracking system is to locate a predefined target object on every frame of a video sequence. Automatic systems span a wide range of applications, such as traffic monitoring (Kamijo et al 2000;Hsieh et al 2006), surveillance (Haritaoglu et al 2000;Collins et al 2001), video retrieval and summarization (Luo et al 2003), vehicle navigation (Hashima et al 1997;Fraundorfer et al 2007), driver assistance (Handmann et al 1998;Avidan 2004), human computer interaction (Wren et al 1997;Liwicki and Everingham 2009) and face analysis (Gunes and Pantic 2010;Cohn et al 1999). Many tracking algorithms indicate that an adaptive approach based on online learning is advantageous to fixed appearance models learned offline (Babenko et al 2011;Ross et al 2008;Mei and Ling 2009).…”
Principal Component Analysis (PCA) is perhaps the most prominent learning tool for dimensionality reduction in pattern recognition and computer vision. However, the 2 -norm employed by standard PCA is not robust to outliers. In this paper, we propose a kernel PCA method for fast and robust PCA, which we call Euler-PCA (e-PCA). In particular, our algorithm utilizes a robust dissimilarity measure based on the Euler representation of complex numbers. We show that Euler-PCA retains PCA's desirable properties while suppressing outliers. Moreover, we formulate Euler-PCA in an incremental learning framework which allows for efficient computation. In our experiments we apply Euler-PCA to three different computer vision applications for which our method performs comparably with other stateof-the-art approaches.
Electronic supplementary material The online version of this article
“…Object tracking is an important task in many computer vision applications such as driver assistance (Handmann et al, 1998;Avidan et al, 2001), video surveillance (Aggarwal and Cai, 1999;Gavrila, 1999;Kettnaker and Zabih, 1999), object-based video compression (Lee et al, 1997;Bue et al, 2002). Various methods have been proposed and improved, from simple and rigid object tracking under a condition of a static camera, to complex and non-rigid object tracking under a condition of a moving camera.…”
“…It has been widely applied in applications such as video surveillance [1], perceptual user interface [2], video coding [3] and driver assistance [4].…”
The Continuously Adaptive Mean Shift algorithm (CAMShift) is an adaptation of mean shift algorithm for object tracking especially for head and face tracking. Traditional CAMShift can not deal with multi-colored object tracking and situations when similar colors exist nearby. In this paper, a new approach towards these problems using CAMShift with weighted back projection is proposed. In our approach, multidimensional histogram with thresholding strategy is utilized.
And a new back projection weighting strategy is proposed for situations when similar colors exist near the tracked object.Through experiments, the results show that the proposed method exceeds the traditional CAMShift in situations with multi-colored object or similar-colored background while keeping the processing speed real-time.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.