Optimized Anfis Model with Hybrid Metaheuristic Algorithms for Facial Emotion Recognition

Emotion recognition from facial images is an important and active area of research. Facial features are widely used in computer vision for emotion interpretation, cognitive science, and social interaction. To obtain accurate analysis of facial expressions (happy, angry, sad, surprised, disgusted, fearful, and neutral), a complex method based on human-computer interaction and data is required. It is still difficult to develop an effective and computationally simple mechanism for feature selection and emotion classification. In this paper, an emotion recognition model using adaptive neuro-fuzzy inference system optimized with particle swarm optimization is proposed. The proposed model was compared with many classification algorithms (ANNs, SVMs, and k-Nearest Neighbor (k-NN) and their subcomponents). The confusion matrix was used to evaluate the performance of these classifiers. The proposed model was evaluated using the MUG database. The model achieved a prediction accuracy of 99.6%.

Automated summary

Emotion recognition from facial images is an important research area. This paper proposes an emotion recognition model based on adaptive neuro-fuzzy inference system optimized with particle swarm optimization, achieving a prediction accuracy of 99.6% on the MUG database.