An effective Face Emotion Recognition using Augmented Physics Informed Neural Networks with Leaf in Wind Optimization

Facial expression recognition plays a crucial role in effective computing and human-computer interaction. This technology enables the automatic detection and interpretation of human emotions based on facial expressions.. There are several uses for facial expression recognition, ranging from bettering human-robot interactions and mental health assessments to increasing customer service and entertainment experiences. In this manuscript, human face emotion recognition using augmented physics-informed neural networks with Leaf in Wind Optimization (HFER-LWO-APINN) is discussed. Initially the input image is collected from FER-2013 Dataset. Then the collected image is fed into pre-processing utilizing Maximum Correntropy Quaternion Kalman Filter (MCQKF). The MCQKF is used to resize and normalize the image then the pre-processed image is fed to Augmented Physics-In formed Neural Networks (APINN) to effectively detect the human emotion and classify the expressions such as angry, disgusted, fearful, sad, happy, surprised and neutral. In general, Augmented Physics-Informed Neural Networks does not express adapting optimization strategies to determine optimal parameters to ensure accurate human emotion detection. Hence, Leaf in Wind Optimization (LWO) is used to enhance the weight parameter of Augmented Physics-Informed Neural Network to accurately detect human emotion and classify the expressions. The effectiveness of the proposed HFER-LWO-APINN method is implemented in Python and evaluated through performance metrics like, Accuracy, Precision, Recall, F1-score, R-square and Mean Absolute Error (MAE).