Least Soft-thresold Squares Tracking

we propose a generative tracking method based on a novel robust linear regression algorithm. In con- trast to existing methods, the proposed Least Soft-thresold Squares (LSS) algorithm models the error term with the Gaussian-Laplacian distribution, which can be solved ef- ficiently. Based on maximum joint likelihood of parameters, we derive a LSS distance to measure the difference between an observation sample and the dictionary. Compared with the distance derived from ordinary least squares methods, the proposed metric is more effective

LSST-MatlabCode-V1
Data
demo_LSS_PCA.m
Evaluation
ReadMe.txt
Trackers
trackparam.m