Learning to track 3D human motion from silhouettes
ICML '04 Proceedings of the twenty-first international conference on Machine learning
Recovering 3D Human Pose from Monocular Images
IEEE Transactions on Pattern Analysis and Machine Intelligence
Spatio-Temporal Context for Robust Multitarget Tracking
IEEE Transactions on Pattern Analysis and Machine Intelligence
Estimating 3D hand pose using hierarchical multi-label classification
Image and Vision Computing
Incremental Learning for Robust Visual Tracking
International Journal of Computer Vision
Posterior probability measure for image matching
Pattern Recognition
Pose estimation and tracking using multivariate regression
Pattern Recognition Letters
Unusual human behavior recognition using evolutionary technique
Computers and Industrial Engineering
Efficient illumination independent appearance-based face tracking
Image and Vision Computing
Object Tracking by Maximizing Classification Score of Detector Based on Rectangle Features
IEICE - Transactions on Information and Systems
Learning AAM fitting through simulation
Pattern Recognition
RVM-based human action classification in crowd through projection and star skeletonization
Journal on Image and Video Processing - Special issue on video-based modeling, analysis, and recognition of human motion
A wavelet subspace method for real-time face tracking
Real-Time Imaging
Learning-based object tracking using boosted features and appearance-adaptive models
ACIVS'07 Proceedings of the 9th international conference on Advanced concepts for intelligent vision systems
3D human pose from silhouettes by relevance vector regression
CVPR'04 Proceedings of the 2004 IEEE computer society conference on Computer vision and pattern recognition
Vehicle trajectory estimation using spatio-temporal MCMC
EURASIP Journal on Advances in Signal Processing - Special issue on advanced image processing for defense and security applications
Real-time tracking with classifiers
WDV'05/WDV'06/ICCV'05/ECCV'06 Proceedings of the 2005/2006 international conference on Dynamical vision
Linear Regression and Adaptive Appearance Models for Fast Simultaneous Modelling and Tracking
International Journal of Computer Vision
Online feature selection using mutual information for real-time multi-view object tracking
AMFG'05 Proceedings of the Second international conference on Analysis and Modelling of Faces and Gestures
Multivariate relevance vector machines for tracking
ECCV'06 Proceedings of the 9th European conference on Computer Vision - Volume Part III
Directional eigentemplate learning for sparse template tracker
PSIVT'11 Proceedings of the 5th Pacific Rim conference on Advances in Image and Video Technology - Volume Part II
A fast approach for person detection and tracking
International Journal of Computer Applications in Technology
Markov Chain Monte Carlo Modular Ensemble Tracking
Image and Vision Computing
| Hi-index | 0.00 |
This paper addresses the problem of applying powerful patternrecognition algorithms based on kernels to efficient visualtracking. Recently Avidan [1] has shown that object recognizersusing kernel-SVMs can be elegantly adapted to localization by meansof spatial perturbation of the SVM, using optic flow. WhereasAvidan's SVM applies to each frame of a video independently ofother frames, the benefits of temporal fusion of data are wellknown. This issue is addressed here by using a fully probabilistic'Relevance Vector Machine' (RVM) to generate observations withGaussian distributions that can be fused over time. To improveperformance further, rather than adapting a recognizer, webuild alocalizer directly using the regression form of the RVM. Aclassification SVM is used in tandem, for object verification, andthis provides the capability of automatic initialization andrecovery. The approach is demonstrated in real-time face andvehicle tracking systems. The 'sparsity' of the RVMs means thatonly a fraction of CPU time is required to track at frame rate.Tracker output is demonstrated in a camera management task in whichzoom and pan are controlled inresponse to speaker/vehicle positionand orientation, over an extended period. The advantages oftemporal fusion inthis system are demonstrated.