Quantum Machine Learning for Facial Expression Recognition

The seamless integration of humans and computers in artificial intelligence endeavors is pivotal in modern society, where digital imaging permeates all aspects of life. Recognizing emotions from facial expressions is a significant area of research, with algorithms leveraging techniques like Local Binary Patterns (LBP) and Histogram of Oriented Gradients (HOG) for real-time analysis. Principal Component Analysis (PCA) further enhances facial recognition systems' efficiency. This study compares a conventional Support Vector Machine (SVM) with a quantum-based SVM (Qsvm) for emotion recognition. Leveraging the power of quantum computing through Qiskit, our Qsvm algorithm aims to surpass classical SVM limitations, particularly in handling large datasets while ensuring data privacy. Rigorous experimentation evaluates the algorithms' performance on standardized datasets, highlighting Qsvm's computational efficiency and potential scalability advantages. Additionally, we address data privacy concerns by imprinting facial expression data into quantum states, ensuring protection from unauthorized access during processing, thus enhancing privacy in sensitive applications.