SPEECH RECOGNITION BASED ON FUSION OF VISUAL AND AUDITORY INFORMATIONUSING FULL-FRAME COLOR IMAGE

Authors
IGAWA S OGIHARA A SHINTANI A TAKAMATSU S
Citation
S. Igawa et al., SPEECH RECOGNITION BASED ON FUSION OF VISUAL AND AUDITORY INFORMATIONUSING FULL-FRAME COLOR IMAGE, IEICE transactions on fundamentals of electronics, communications and computer science, E79A(11), 1996, pp. 1836-1840
Citations number
3
Categorie Soggetti
Engineering, Eletrical & Electronic","Computer Science Hardware & Architecture","Computer Science Information Systems
Journal title
IEICE transactions on fundamentals of electronics, communications and computer science → ACNP
ISSN journal
09168508
Volume
E79A
Issue
11
Year of publication
1996
Pages
1836 - 1840
Database
ISI
SICI code
0916-8508(1996)E79A:11<1836:SRBOFO>2.0.ZU;2-B
Abstract
We propose a method to fuse auditory information and visual informatio n for accurate speech recognition. This method fuses two kinds of info rmation by using linear combination after calculating two kinds of pro babilities by HMM for each word. In addition, we use full-frame color image as visual information in order to improve the accuracy of the pr oposed speech recognition system. We have performed experiments compar ing the proposed method with the method using either auditory informat ion of visual information, and confirmed the validity of the proposed method.