Output details
15 - General Engineering
University of Hull
Article title
Small-vocabulary speech recognition using a silent speech interface based on magnetic sensing
Type
D - Journal article
Title of journal
Speech Communication
Article number
-
Volume number
55
Issue number
1
First page of article
22
ISSN of journal
0167-6393
Year of publication
2013
Number of additional authors
6
Additional information
This paper presents the first demonstration of the recognition of connected sequences of words using the PMA technique. This is a very important step towards reconstruction of normal, fluent speech and, at 96% word accuracy, recognition performance is shown to be better than many audio-based recognition systems. An early version of this paper was selected as one of the best papers (top 2.5%) from the Interspeech-2010 conference and selected for journal publication. The unique processing algorithm developed to reconstruct the speech is currently being patented. Preparations are now underway to start a clinical trial using the system in 2014.
Interdisciplinary
-
Cross-referral requested
-
Research group
None
Proposed double-weighted
No
Double-weighted statement
-
Reserve for a double-weighted output
No
Non-English
No
English abstract
-
