—Discriminative Training (DT) methods for acoustic modeling, such as MMI, MCE, and SVM, have been proved effective in speaker recognition. In this paper we propose a DT method fo...
Estimation based on received signal strength (RSS) is crucial in sensor networks for sensor localization, target tracking, etc. In this paper, we present a Gaussian approximation ...
Volkan Cevher, Aswin C. Sankaranarayanan, Rama Che...
We present a Bayesian approach to color constancy which utilizes a nonGaussian probabilistic model of the image formation process. The parameters of this model are estimated direc...
Charles R. Rosenberg, Thomas P. Minka, Alok Ladsar...
—We consider distributed estimation of the inverse covariance matrix in Gaussian graphical models. These models factorize the multivariate distribution and allow for efficient d...
Maximum entropy models are a common modeling technique, but prone to overfitting. We show that using an exponential distribution as a prior leads to bounded absolute discounting b...