The use of the IBM 704 in the simulation of speech-recognition systems

The term speech-recognition device usually brings to mind a machine capable of duplicating the function of a human listener. Such a device not only would have to be capable of receiving and classifying acoustic stimuli, but also would have to be able to extract from these stimuli the message the speaker intended. To do this, a recognition machine must be familiar with the language statistics, and indeed, the entire human environment. Many years of investigation will be required before such a human replacement can be achieved.