Quantum K-Nearest Neighbors for Emotion Recognition Using Visual Art Based on Color and Texture Analysis

The emotion recognition is the process of recognizing a person emotion in various ways that also expressed in arts such as paintings, drawings are the visual art. In traditional automated detection methods performs based on fixed rules and manual steps which affect to recognize the arts. To overcome this problem, Quantum K-Nearest Neighbors (QKNN) for used classify the emotions using visual art. The WikiArt and AffectNet are collection of images, artworks based on visual art that provide as input to the proposed method QKNN. The gray scale conversion and data augmentation is used for pre-processing for improve the quality of images. The Histogram of Oriented Gradients (HOG) is used for extract features of pre-processed data. The KNN is used for classify the emotions and improved QKNN is utilized for enhance the accuracy of the classification. The performance metrices are used to evaluate the overall results of metrics such as accuracy of 98.46% and 99.12% for WikiArt and AffectNet are compared with existing works of Support Vector Machine (SVM) and Convolutional Neural Network (CNN) respectively.