Open AccessAutoAssociative Memory for Face Recognition using High-Dimensional Octonion Data
Author(s) -
Shoba Rani Salvadi,
D. Nagendra Rao,
Srinivasan Vatsal
Publication year - 2021
Publication title -
iop conference series. materials science and engineering
Language(s) - English
Resource type - Journals
eISSN - 1757-899X
pISSN - 1757-8981
DOI - 10.1088/1757-899x/1042/1/012028
Subject(s) - quaternion , generalization , face (sociological concept) , artificial intelligence , computer science , facial recognition system , content addressable memory , hypercomplex number , computer vision , representation (politics) , identity (music) , uniqueness , pattern recognition (psychology) , algebra over a field , mathematics , artificial neural network , pure mathematics , geometry , aesthetics , mathematical analysis , social science , philosophy , sociology , politics , political science , law
Uniqueness or individuality of an individual face is the representation of one’s identity. Recent research in color image processing focusing on octonion algebra, which is the generalization of complex and quaternion algebra. In this paper, the face of an individual is used for identification making automatically. For this, an autoassociative memory is used for the recognition of images or videos with 8-dimensions. Octonion algebra is not commutative and not associative but it is an alternate algebra for representing the phase and angle of images effectively. The prospective experiments are conducted on color images for image reconstruction, denoising and efficiently retrieving blurred images in true color images.