Browse Prior Art Database

Polar Coordination Display for Speech Recognition Products

IP.com Disclosure Number: IPCOM000104927D
Original Publication Date: 1993-Jun-01
Included in the Prior Art Database: 2005-Mar-19
Document File: 2 page(s) / 56K

Publishing Venue

IBM

Related People

Crepy, H: AUTHOR [+2]

Abstract

Described is an hardware implementation utilizing polar coordinates to display word matching in speech recognition applications. The implementation uses the graphical display of the results of word recognition providing feedback that can be used to improve word pronunciation. The implementation is particularly useful in speech therapy applications.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 52% of the total text.

Polar Coordination Display for Speech Recognition Products

      Described is an hardware implementation utilizing polar
coordinates to display word matching in speech recognition
applications.  The implementation uses the graphical display of the
results of word recognition providing feedback that can be used to
improve word pronunciation.  The implementation is particularly
useful in speech therapy applications.

      The implementation is an improved speech recognition method to
provide a visually attractive and an easily read display of the
accuracy of word matching by using the position of maximum difference
between the utterance and its model in a word recognizer.  Utilizing
a word recognition system for speech therapy generally requires that
the results of the recognition be presented in such a way that the
information can be used as feedback to correct the pronunciation.
The simple indication of a match/no match approach, as is usually
provided in word recognizer devices, is not considered sufficient for
this purpose.  This is because the display of the results is not
visually appealing nor is the device compact and portable.

      The concept described herein is designed to provide a word
recognizer output that measures the distance from the utterance to
the model word and to provide a measure of when in the utterance the
maximum deviation from the model word was observed.  Recognition of
incoming words are recognized against models of one to fou...