Face and Speech Recognition Fusion in Personal Identification

Security personnel manning access points most often based their access authorisation on recognition of faces. And it is also very common to base access decision for a person knocking at the door on recognition of his/her voice. The conventional manual method of drawing attention by knocking the door...

Full description

Saved in:
Bibliographic Details
Published in:International journal of computer applications 2012-01, Vol.47 (23), p.36-41
Main Authors: S, Ibiyemi T, A, Aliu S, G, Akintola A
Format: Article
Language:English
Subjects:
Commands
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Security personnel manning access points most often based their access authorisation on recognition of faces. And it is also very common to base access decision for a person knocking at the door on recognition of his/her voice. The conventional manual method of drawing attention by knocking the door, pressing door bell, or/and shouting one's presence in order to gain access are inefficient and risky. A better method is automatic personal identification based on face and speech recognition which is the subject of this paper. Eigenface method is used for the face recognition. While the speaker and spoken command recognition are both based on the same mel-frequency cepstral coefficients as feature vectors extracted from English and Yorùbá utterances. Experiments yielded 90% face recognition while recognition rates for speakers and the spoken commands were 87% and 74% for utterances in English and Yorùbá respectively. However, the final recognition decision for authorisation or access activation is based on the recognition outcomes of the face recognition and the speech recognition.
ISSN:0975-8887
0975-8887
DOI:10.5120/7498-0617