An Efficient Deep Learning Technique for Facial Expression Recognition

Automatic facial expression recognition is one of the most important issues in computer vision and it has its use in areas such as HCI and emotion-based systems. For facial expression recognition, this paper presented a Bidirectional Long Short-Term Memory (BiLSTM) deep learning technology. The proposed model is to address the temporal structure of sequential facial features and provided improved performance in the temporal dependency of facial expressions. We utilize a selected dataset, perform preprocessing on the facial images, BacYN feature extraction, and optimize BiLSTM for expression recognition. Based on a large number of experiments and assessments, our method proves successful and can be successfully used in realizing applications that require a detailed analysis of facial emotions.