—We propose a general methodology (PURE-LET) to design and optimize a wide class of transform-domain thresholding algorithms for denoising images corrupted by mixed Poisson–Gau...
This paper investigates the performance of machine learning methods for classifying rock types from hyperspectral data. The main objective is to test the impact on classification ...
This paper describes an adaptive noise cancellation structure suitable for the separation and equalization of signals that coupled due to the approximate placing of primary and au...
In automatic speech recognition, we are faced with a wellknown inconsistency: Bayes decision rule is usually used to minimize sentence (word sequence) error, whereas in practice w...
The best known method to find exact or at least -approximate solutions to polynomial programming problems is the spatial Branch-and-Bound algorithm, which rests on computing lower...