PROJECT TITLE :

Interacting Multiview Tracker

ABSTRACT:

A robust algorithm is proposed for tracking a target object in dynamic conditions including motion blurs, illumination changes, cause variations, and occlusions. To deal with these challenging factors, multiple trackers based on totally different feature representations are integrated within a probabilistic framework. Every view of the proposed multiview (multi-channel) feature learning algorithm is anxious with one explicit feature representation of a target object from which a tracker is developed with different levels of reliability. With the multiple trackers, the proposed algorithm exploits tracker interaction and choice for strong tracking performance. Within the tracker interaction, a transition likelihood matrix is employed to estimate dependencies between trackers. Multiple trackers communicate with each different by sharing data of sample distributions. The tracker choice method determines the most reliable tracker with the highest chance. To account for object look changes, the transition likelihood matrix and tracker chance are updated during a recursive Bayesian framework by reflecting the tracker reliability measured by a sturdy tracker chance function that learns to account for each transient and stable look changes. Experimental results on benchmark datasets demonstrate that the proposed interacting multiview algorithm performs robustly and favorably against state-of-the-art strategies in terms of many quantitative metrics.