Improvement of Speech Preprocessing Model Based on Linear Prediction

Abstract

With the rapid development of information society and communication technology, human-computer interaction becomes more and more important. Ther1 efore, the recognition rate of speech signal becomes the target that people pursue. Speech preprocessing plays an important role in achieving this goal. In this paper,we firstly analyze the conventional linear prediction model and propose a new speech signal regression linear prediction model based on the linear prediction model algorithm, Type by regression forecasting model to forecast the speech signal processing, can more quickly identity, by the MATLAB simulation results show that the new type regression forecasting model performance than conventional linear prediction model is nearly 200 ms of acceleration performance, and greater effectiveness of improved prediction model, namely the pitch detection in the detection performance of speech waveform change processing, greatly improves the speech recognition rate.