Segmenting the speech signals on the basis of time-frequency analysis is the most natural approach. Boundaries are located in places where energy of some frequency subband rapidly...
In this paper the potential of GP-generated symbolic regression for alleviating multicollinearity problems in multiple regression is presented with a case study in an industrial s...
Abstract. We have carried out a detailed evaluation of the use of texture features in a query-by-example approach to image retrieval. We used 3 radically different texture feature...
We present a data-driven approach to predict the importance of edges and construct a Markov network for image analysis based on statistical models of global and local image feature...
We propose a general framework for performing independent component analysis (ICA) which relies on ensemble learning and linear response theory known from statistical physics. We ...