Dec 31, 2010

Automated speech processing: commercial state of the art



I have always been fascinated by progress in the area of automated speech processing, since my days as a Master's student at the University of Cambridge in Computer Speech and Language Processing and I spent time with Professor Steve Young at the Cambridge University Machine Intelligence Laboratory.

Thus it was with great interest and pleasure that I read this review of the state of the art of commercial speech processing systems within Scientific American earlier this month. Like many high tech fields, there have been mergers and convergence in the suppliers, to the leader Nuance Communications. Extensive pre-training is no longer required. General listening is now possible in structured sub-languages and categories. Plus, the commercial applications have become widespread, especially related to phone interfaces.

Why is this relevant for healthcare relationship marketing. Possible applications include:

* Acquisition and registration into programs via mobile smartphones, for either patients or professionals.

* Automated dictation transcription processing of the highly structured clinical reports that physicians issue after seeing patients.

* speech to text social network inputs as a convenience, or for the visually impaired.

* processing call center transcripts for marketing intelligence.

* insert your brainstorm here.