Paper Info
Title
Online Multiple Instance Joint Model for Visual Tracking
Abstract
Although numerous online learning strategies have beenproposed to handle the appearance variation in visualtracking, the existing methods just perform well in certaincases since they lack effective appearance learning mechanism. In this paper, a joint model tracker (JMT) is presented, which consists of a generative model based on MultipleSubspaces and a discriminative model based on improvedMultiple Instance Boosting (MIBoosting). The generativemodel utilizes a series of local constructed subspacesto update the Multiple Subspaces model and considersthe energy dissipation of dimension reduction in updatingstep. The discriminative model adopts the GaussianMixture Model (GMM) to estimate the posterior probabilityof the likelihood function. These two parts supervise eachother to update in multiple instance way which helps ourtracker recover from drift. Extensive experiments on variousdatabases validate the effectiveness of our proposedmethod over other state-of-the-art trackers.
Year
DOI
Venue
2012
10.1109/AVSS.2012.52
AVSS
Keywords
Field
DocType
considersthe energy dissipation,gaussianmixture model,discriminative model,visual tracking,dimension reduction,multiple subspaces model,online multiple instance joint,effective appearance,existing method,appearance variation,joint model tracker,generative model,gaussian mixture model,learning artificial intelligence,computational modeling,covariance matrix,mathematical model,gaussian processes,object tracking,maximum likelihood estimation,boosting
Dimensionality reduction,Likelihood function,Pattern recognition,Computer science,Video tracking,Eye tracking,Artificial intelligence,Boosting (machine learning),Discriminative model,Machine learning,Mixture model,Generative model
Conference
Citations 
PageRank 
References 
4
0.40
17
Authors
6
Name
Order
Citations
PageRank
Longyin Wen164733.89
Zhaowei Cai245216.64
Menglong Yang310910.49
Zhen Lei43613157.95
Dong Yi5117343.66
Stan Z. Li68951535.26