A global robust M-estimation scheme for maximum a posteriori (MAP) image super-resolution which efficiently addresses the presence of outliers in the low-resolution images is proposed. In iterative MAP image super-resolution, the objective function to be minimized involves the highly resolved image, a parameter controlling the step size of the iterative algorithm, and a parameter weighing the d...

In the past few years, transformation-based model adaptation techniques have been widely used to help reducing acoustic mismatch between training and testing conditions of automatic speech recognizers. The estimation of the transformation parameters is usually carried out using estimation paradigms based on classical statistics such as maximum likelihood, mainly because of their conceptual and ...

The context-tree weighting method (Willems, Shtarkov, and Tjalkens [1995]) can be used to compress sequences generated by tree sources. Its redundancy behavior is optimal in the sense that Rissanen’s lower bound [1984] is achieved. Here we study some questions related to the context-tree weighting method. First we stress again that the a priori distribution over all tree models that is mainly c...

Perturbation models are families of distributions induced from perturbations. They combine randomization of the parameters with maximization to draw unbiased samples. Unlike Gibbs’ distributions, a perturbation model defined on the basis of low order statistics still gives rise to high order dependencies. In this paper, we analyze, extend and seek to estimate such dependencies from data. In par...

We consider the problem of estimating a pattern of faults, represented as a binary vector, from a set of measurements. The measurements can be noise corrupted real values, or quantized versions of noise corrupted signals, including even 1-bit (sign) measurements. Maximum a posteriori probability (MAP) estimation of the fault pattern leads to a difficult combinatorial optimization problem, so we...

A widely used method to determine the a priori SNR from noisy speech is the decision directed (DD) approach, but the a priori SNR follows the a posteriori SNR with a delay of one frame in speech frames. As a consequence, the performance of the noise reduction system degrades. In order to overcome this artifact, we propose three computationally simple and efficient two-step methods based on the ...

The process of turbo-code decoding starts with the formation of a posteriori probabilities (APPs) for each data bit, which is followed by choosing the data-bit value that corresponds to the maximum a posteriori (MAP) probability for that data bit. Upon reception of a corrupted code-bit sequence, the process of decisionmaking with APPs allows the MAP algorithm to determine the most likely inform...

This paper studies the problem of simultaneous input and state estimation (SISE) for nonlinear dynamical systems with and without direct input-output feedthrough. We take a Bayesian perspective to develop a sequential joint input and state estimation approach. Our scheme gives rise to a nonlinear Maximum a Posteriori optimization problem, which we solve using a classical Gauss-Newton method. Th...

In this paper, we propose a novel adaptation technique based on coarse/fine training of transfer vectors. We focus on transfer vector estimation of a Gaussian mean from an initial model to an adapted model. The transfer vector is decomposed into a direction vector and a scaling factor. By using tied-Gaussian class (coarse class) estimation for the direction vector, and by using individual Gauss...

In this paper we present a finite dimensional iterative algorithm for optimal maximum a posteriori (MAP) state estimation of bilinear systems. Bilinear models are appealing in their ability to represent or approximate a broad class of nonlinear systems. We show that several bilinear models previously considered in the literature are special cases of the general bilinear model we propose. Our it...