Method of independently creating and using a garbage model for improved rejection in a limited-training speaker-dependent speech recognition system

Abstract

A speaker-dependent (SD) speech recognition system. The invention is specifically tailored to operate with very little training data, and also within hardware constraints such as limited memory and processing resources. A garbage model and a vocabulary model are generated and are subsequently used to perform comparison to a speech signal to decide if the speech signal is a specific vocabulary word. A word score is generated, and it is compared to a number of parameters, including an absolute threshold and another word score. Off-line training of the system is performed, in one embodiment, using compressed training tokens. A speech signal is segmented into scramble frames wherein the scramble frames have certain characteristics. For example, length is one characteristic of the scramble frames, each scramble frame having a length of an average vowel sound, or a predetermined length of nominally 40-50 msec. The invention is operable to be trained using as little as one single training token that is segmented. Those segments may be re-arranged to form a pseudo-token to form a garbage model. The use of a pseudo-token allows for generation of a reliable garbage model having many speaker-specific characteristics of an original training token while discarding the specific acoustic characteristics of any vocabulary word corresponding to the training token. The invention is equally as operable by using a training token to form a vocabulary model having multiple states, and re-arranging those states to form one or more garbage models.