Intelligent Quantum Machine Learning Framework for Emotion-Aware Cyber-Physical Systems Supporting Children with Down Syndrome
Year: 2026
Type: conference-paper
Abstract: Many children with Down syndrome have a hard time understanding and putting their feelings into words. This can make it tougher for them to connect with others and to fully take part in school and community life. To help address this, the study developed an Intelligent Quantum Machine Learning Framework for Emotion-Aware Cyber-Physical Systems designed to support children with Mild Down Syndrome. The framework brought together Affective Computing, Quantum Machine Learning (QML), Cyber-Physical Systems (CPS), Human-Centered Design (HCD), and the Ikigai Framework to build an intelligent system capable of recognizing emotions and providing real-time emotional awareness. Guided by the System Development Life Cycle (SDLC) methodology, the research gathered emotional data through facial expressions, eye-gaze patterns, and behavioral cues captured by sensing devices. These inputs were processed using affective computing techniques and classified with three Quantum Machine Learning models - the Variational Quantum Classifier (VQC), Quantum Support Vector Machine (QSVM), and Quantum Neural Network (QNN) - simulated through IBM Qiskit. The system was built to identify six emotional states: Happy, Sad, Angry, Fear, Surprise, and Neutral. The Quantum Neural Network delivered the strongest results, reaching 95.83% accuracy. Overall, the framework achieved 95.51% accuracy, 95.29% precision, 95.51% recall, and a 95.39% F1-score, with an average classification time of 42.35 ms and response time of 65.93 ms while using only 8 qubits and a circuit depth of 18. These findings show that combining Quantum Machine Learning with Affective Computing and Cyber-Physical Systems can meaningfully strengthen emotional awareness for children with Mild Down Syndrome, offering teachers, therapists, caregivers, and families timely insights that support communication, social interaction, and development.
Language: English
Topic: Emotion and Mood Recognition
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7220720149