EEG-Based Emotion Recognition via Knowledge-Integrated Interpretable Method

Despite achieving success in many domains, deep learning models remain mostly black boxes, especially electroencephalogram (EEG)-related tasks. Meanwhile, understanding the reasons behind model predictions is quite crucial assessing trust and performance promotion EEG-related In this work, we explore use of representative interpretable to analyze behavior convolutional neural networks (CNN) EEG-based emotion recognition. According analysis, find that similar features captured by our state-of-the-art are consistent with previous brain science findings. Next, propose a new integrating knowledge interpretability analysis results process. Our knowledge-integrated achieves better recognition accuracy on standard datasets.