Stuart N WrigleyBSc(Hons) PhD MIET SMIEEE
Operations and Business Development Manager, UKRI CDT in Speech and Language Technologies and their Applications

Speech Recognition by Dynamic Time Warping

Automatic speech recognition (ASR) is an important field of research and has applications in many areas. A well-studied approach to automatic speech recognition is based on the storage of one or more acoustic patterns (templates) for each word in the recognition vocabulary. The recognition process then consists of matching the incoming speech with stored templates. The template with the lowest distance measure from the input pattern is the recognised word. One algorithm used to find the best match (lowest distance measure) is based upon dynamic programming. The aim of this project was to implement a Dynamic Time Warping (DTW) word recogniser.

The pages are structured into separate topic areas which can be explored using the second menu at the top of this page.

First created on 20 April 1998 by Stuart N Wrigley (s.wrigley@dcs.shef.ac.uk)