Facial emotion recognition using deep quantum and advanced transfer learning mechanism.

Area of Science:

• Computer Science
• Artificial Intelligence
• Quantum Computing

Background:

• Facial expression recognition is crucial for human interaction and emotion analysis.
• Traditional methods struggle with accurately predicting emotions from visual cues.
• Deep Learning (DL) algorithms are increasingly used for facial expression detection.

Purpose of the Study:

• To propose a novel system for facial expression recognition.
• To leverage quantum computing for enhanced feature extraction and reduced computation time.
• To improve the accuracy and efficiency of emotion analysis.

Main Methods:

• Utilized a Modified ResNet model with residual connections.
• Employed quantum computing for feature extraction via a quantum convolutional layer.
• Integrated the Modified up Sampled Bottle Neck Process (MuS-BNP) with ResNet-18.

Main Results:

• Achieved superior performance in differentiating facial expressions.
• Demonstrated significant reduction in computation time compared to conventional methods.
• Validated through metrics including accuracy, F1-score, recall, and precision.

Conclusions:

• The proposed system demonstrates quantum superiority in facial expression recognition.
• Quantum computing integration offers faster and more accurate computations.
• This approach presents a promising solution for advanced emotion analysis tasks.