Emotion Recognition Based on Electroencephalography: survey

Abstract

Electroencephalography (EEG) is considered as a standard method for recording human brain activity in the form of electric pulses over time. The significant attention in emotion recognition based on EEG has grown meaningfully, owing to its potential applications in brain-computer interfaces, mental health monitoring, and human-computer interaction. This article provided a comprehensive literature review of emotion recognition using EEG data, investigating datasets, classification techniques, challenges, and real-world applications. Various publicly available EEG datasets, including Database for Emotion Analysis using Physiological Signals (DEAP), Shanghai Jiao Tong University Emotion EEG Dataset (SEED), and Multimodal Dataset for Emotion Recognition (DREAMER), are investigated in terms of their structure, features, and labeling techniques. Additionally, we also investigated various classification methods, from conventional machine learning algorithms like support vector machines (SVM) to sophisticated deep learning models such as convolutional neural networks (CNN) and recurrent neural networks (RNN). Even with significant advances, issues like inter-subject variability, data noise, and the necessity for reliable feature extraction techniques continue to pose major obstacles to attaining improved recognition accuracy. Furthermore, the review underscored the expanding use of EEG-based emotion recognition in fields such as affective computing, neurofeedback systems, and adaptive learning environments. This study offered valuable insights into future research directions for enhancing EEG-based emotion recognition systems by synthesizing recent advancements and challenges.  Finally, our findings can be summarized as: mapping the current research landscape, evaluating the effectiveness of tools and techniques, identifying challenges and limitations, and recommending future paths for researchers and developers.