Intelligent preprocessing and classification of audio signals

Abstract

An audio processor that integrates intelligent classification and preprocessing algorithms is presented. Audio features in the time and frequency domains are extracted and processed prior to classification. Classification algorithms, including the nearest neighbor rule (NNR), artificial neural networks (ANN), fuzzy neural networks (FNN), and hidden Markov models (HMM), are used to classify and identify singers and musical instruments. A training phase is required to establish a feature space template, followed by a test phase in which the audio features of the test data are calculated and matched to the feature space template. In addition to audio classification, the proposed system provides several independent component analysis (ICA)-based preprocessing functions for blind source separation, voice removal, and noise reduction. The proposed techniques were applied to process various kinds of audio program materials. The test results reveal that the performance of the methods is satisfactory, but varies slightly with the algorithm and program materials used in the tests.