期刊名称：International Journal of Computer Science and Information Technologies
电子版ISSN：0975-9646
出版年度：2011
卷号：2
期号：4
页码：1555-1558
出版社：TechScience Publications
摘要：Pattern Recognition is the process to classify data or patterns based on either a priori knowledge or on statistical information extracted from the patterns. An audio pattern recognition problem is based on speech patterns spoken, which can be interpreted as speaker dependent or speaker independent. Artificial Neural Network (ANN) is information processing machine learning model, inspired by biological neural systems. ANN has a potential of massive computing, online adaptation and learning abilities. Neural network consists of many simple processing elements joined by weighted connection paths. A neural net produces an output signal in response to an input pattern; the output is determined by value of weights. This paper discusses various neural network paradigms for audio pattern recognition and based on the study a new paradigm for audio pattern recognition is suggested.
