Konuşma Tanıma için İnsan-makine Karşılaştırması

Speech/voice recognition by machines has been a topic of interest since 1950s. Research that initially adopted dynamic programming methodologies now mostly uses the hidden Markov model as the method for speech recognition. Nevertheless, even the most advanced speech recognition system makes, depending on the context, 2-20 times more errors than humans. Although the basic principles behind human speech recognition have not been completely understood, there are some theories that attempt to explain biological mechanisms for speech recognition. This paper aims to provide a review of these theories as well as a brief history of developments in automatic speech recognition technology. Furthermore, the paper discusses some recent studies on Turkish speech recognition. The paper concludes with a comparison between human and machine speech recognition performance