Exploration of Subjective Color Perceptual-Ability by EEG-Induced Type-2 Fuzzy Classifiers

• Published in: IEEE transactions on cognitive and developmental systems Vol. 12; no. 3; pp. 618 - 635
• Main Authors: Laha, Mousumi, Konar, Amit, Rakshit, Pratyusha, Nagar, Atulya K.
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 01.09.2020; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Brain; Classifiers; Color; Color perceptual-ability; Colored noise; Electroencephalography; exact low-resolution electromagnetic topographic analysis (e-LORETA); Feature extraction; Fuzzy logic; Fuzzy sets; Image color analysis; Noise tolerance; Statistical tests; type-2 fuzzy classifiers; type-2 fuzzy sets; Uncertainty
• ISSN: 2379-8920; 2379-8939
• DOI: 10.1109/TCDS.2019.2959138

Perceptual-ability informally refers to the ability of a person to recognize a stimulus. This article deals with color perceptual-ability measurement of subjects using brain response to basic color (red, green, and blue) stimuli. It also attempts to determine the subjective ability to recognize the base colors in the presence of noise tolerance of the base colors, referred to as recognition tolerance. Because of intrasession and intersession variations in subjective brain signal features for a given color stimulus, there exists uncertainty in perceptual-ability. In addition, small variations in the color stimulus result in wide variations in brain signal features, introducing uncertainty in the perceptual-ability of the subject. Type-2 fuzzy logic has been employed to handle the uncertainty in color perceptual-ability measurements due to: 1) variations in brain signal features for a given color and 2) the presence of colored noise on the base colors. Because of the limited power of uncertainty management of the interval type-2 fuzzy sets and high computational overhead of its general type-2 counterpart, we develop a semi-general type-2 fuzzy classifier to recognize the base color. It is important to note that the proposed technique transforms a vertical slice-based general type-2 fuzzy set (GT2FS) into an equivalent interval type-2 counterpart to reduce the computational overhead, without losing the contributions of the secondary memberships. The proposed Semi-GT2FSs-induced classifier yields superior performance in classification accuracy with respect to existing type-1, type-2, and other well-known classifiers. The brain-understanding of a perceived base or noisy base colors is also obtained by exact low-resolution electromagnetic topographic analysis (e-LORETA) software. This is used as the reference for our experimental results of the semi-general type-2 classifier in color perceptual-ability detection. Statistical tests undertaken confirm the superiority of the proposed classifier over its competitors. The proposed technique is expected to have interesting applications in identifying people with excellent color perceptual-ability for chemical, pharmaceutical, and textile industries.