Visual tracking and recognition using appearance-adaptive models in particle filters

TitleVisual tracking and recognition using appearance-adaptive models in particle filters
Publication TypeJournal Articles
Year of Publication2004
AuthorsZhou S K, Chellappa R, Moghaddam B
JournalIEEE Transactions on Image Processing
Pagination1491 - 1506
Date Published2004/11//
ISBN Number1057-7149
Keywordsadaptive filters, adaptive noise variance, algorithms, appearance-adaptive model, Artificial intelligence, Cluster Analysis, Computer Graphics, Computer simulation, Feedback, Filtering, first-order linear predictor, hidden feature removal, HUMANS, Image Enhancement, Image Interpretation, Computer-Assisted, image recognition, Information Storage and Retrieval, Kinematics, Laboratories, Male, Models, Biological, Models, Statistical, MOTION, Movement, Noise robustness, Numerical Analysis, Computer-Assisted, occlusion analysis, Particle filters, Particle tracking, Pattern Recognition, Automated, Predictive models, Reproducibility of results, robust statistics, Sensitivity and Specificity, Signal Processing, Computer-Assisted, State estimation, statistical analysis, Subtraction Technique, tracking, Training data, visual recognition, visual tracking

We present an approach that incorporates appearance-adaptive models in a particle filter to realize robust visual tracking and recognition algorithms. Tracking needs modeling interframe motion and appearance changes, whereas recognition needs modeling appearance changes between frames and gallery images. In conventional tracking algorithms, the appearance model is either fixed or rapidly changing, and the motion model is simply a random walk with fixed noise variance. Also, the number of particles is typically fixed. All these factors make the visual tracker unstable. To stabilize the tracker, we propose the following modifications: an observation model arising from an adaptive appearance model, an adaptive velocity motion model with adaptive noise variance, and an adaptive number of particles. The adaptive-velocity model is derived using a first-order linear predictor based on the appearance difference between the incoming observation and the previous particle configuration. Occlusion analysis is implemented using robust statistics. Experimental results on tracking visual objects in long outdoor and indoor video sequences demonstrate the effectiveness and robustness of our tracking algorithm. We then perform simultaneous tracking and recognition by embedding them in a particle filter. For recognition purposes, we model the appearance changes between frames and gallery images by constructing the intra- and extrapersonal spaces. Accurate recognition is achieved when confronted by pose and view variations.