Online pattern recognition for Portuguese phonemes using Multi-layer Perceptron combined with recurrent non-linear autoregressive Neural Networks with exogenous inputs
Diana A. Bonilla C, Nadia Nedjah, Luiza de Macedo Mourelle · 2016
Off-line pattern recognition in speech signals is a complex task. Yet, this task becomes harder when the recognition result is required online. The present work proposes an online identification of the Portuguese language phonemes using an nonlinear autoregressive model with exogenous inputs, commonly called NARX. The process first extracts the frequency characteristics of the input speech signals and pre-classifies them into one of the ten possible groups of phonemes, as available in the Portuguese language. This pre-classification is done using a multilayer perceptron network (MLP) with a supervised learning. Subsequently, the MLP output vector, together with the vector that carries the input frequencies, feeds a NARX neural network by means of a temporal delay of four times and feed-backward recurrent links that encompass the results of all hidden layers of the network. As a result of this process, the proposed phoneme recognition process improves the accuracy of the online identification of the Portuguese spoken phonemes during a natural conversation.