Design of an Immersive Experience Emotion Recognition and Analysis System Based on Deep Learning

This paper presents a comprehensive research study on the emotional recognition algorithm for immersive experiences based on deep learning. In recent years, immersive technologies such as virtual reality (VR) and augmented reality (AR) have revolutionized the way people interact with digital content, creating highly engaging and interactive experiences. Understanding users’ emotions during these immersive experiences is becoming increasingly important for various applications, including entertainment, therapy, education, and training. However, accurately detecting emotions from actions in immersive environments remains a challenging task. To address this challenge, this paper proposes a deep learning-based emotional recognition algorithm specifically designed for immersive experiences. Aiming at the shortcoming of the convolutional neural network based on deep and shallow feature fusion in action emotion recognition without considering the temporal sequence of actions, a emotion recognition model based on CNN-BLSTM was designed. Firstly, the action structure features were extracted from the action data using the deep and shallow feature fusion-based action emotion recognition network, integrating features from different convolutional layers. Then, BLSTM extracted temporal dependencies from the fused features, combining CNN and BLSTM to achieve accurate recognition of action emotions. Finally, comparative experiments verified the superior recognition performance of the model on action emotions.