Author: Mengjie Zhang
Source: GZipped PostScript (105kb); Adobe PDF (469kb)
This paper presents a brief overview of speech recognition and machine learning. The basic concepts, dimensions and a common structure of speech recognition and some techniques commonly used in speech recognition are firstly reviewed. After giving a series of concepts, data sets and major paradigms of machine learning, we briefly describe two learning approaches, neural networks and genetic algorithms. Some related work such as how to build a simple speech recognition system, existing HMM software packages and speaker recognition is also presented. We expect that this report can provide background knowledge for people who plan to do or just started doing research and applications in speech recognition using machine learning methods.