A Novel Neural Network Architecture for Facial Emotion Recognition

ABSTRACT Facial Emotion Recognition (FER) has become an important research area in computer vision and artificial intelligence due to its wide range of applications in human–computer interaction, healthcare, e-learning, surveillance, and marketing. Traditional machine learning techniques such as Random Forest and Support Vector Machines rely heavily on handcrafted features, which limits their ability to generalize across complex and diverse datasets. To address these limitations, this paper proposes a deep learning-based FER framework using the MobileNetV2 architecture. MobileNetV2 is a lightweight convolutional neural network designed for mobile and embedded devices while maintaining high accuracy. The proposed model utilizes transfer learning to classify six fundamental human emotions: happiness, sadness, anger, fear, surprise, and disgust. Index Terms: Facial Emotion Recognition, Deep Learning, MobileNetV2, CNN, Transfer Learning, Computer Vision.