Automated Speech Recognition: is Learning Superior to Optimality?

Automated speech recognition, historically posed as a solvable problem, has thus far eluded algorithmic solutions and continues to be outperformed by humans. Conventional speech recognizers employ a training phase during which many of their parameters are configured. During normal operation, these recognizers do not significantly alter these parameters. Conversely the model proposed in this book draws heavily on high level human thought patterns and speech perception to outline a set of precepts to eliminate this training phase and instead opt to perform all its tasks during normal operation. Background on the problems in this field and their classical solutions are presented followed by motivation and implementation details of the proposed model. The testing results of this model indicate that benefits can be seen in increased speech recognizer adaptability while still retaining competitive recognition rates in controlled environments. This book is intended for researchers in artificial intelligence and speech recognition. It is also suitable for for those attempting to learn the subject area.