Scholarlink

EEG signals in emotion recognition is tough due to the difficulty of getting accurate data for emotional states. It can improve brain-computer interface (BCI) systems for therapeutic applications and human social interaction. EEG based emotion classification has evolved into a critical problem in affective computing and cognitive interaction. To improve the accuracy of identifying emotions, it is still difficult to effectively combine the spatial, temporal and spectral distinct details of EEG signals. This study mainly focuses on combining and concatenating the features from SEED dataset such as PSD, DE, RASM, DASM, and ASM for the classification of emotions. LSTM model is then used for performing the classification based on these features.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Deep Learning-Based Emotion Recognition with Multi-feature EEG Fusion Using LSTM

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Ayushi Kotwal,
  • Kush Sharma,
  • Mohsin Haassan,
  • Rahul Sharma,
  • Meetali Verma,
  • Jatinder Manhas,
  • Vinod Sharma

摘要

EEG signals in emotion recognition is tough due to the difficulty of getting accurate data for emotional states. It can improve brain-computer interface (BCI) systems for therapeutic applications and human social interaction. EEG based emotion classification has evolved into a critical problem in affective computing and cognitive interaction. To improve the accuracy of identifying emotions, it is still difficult to effectively combine the spatial, temporal and spectral distinct details of EEG signals. This study mainly focuses on combining and concatenating the features from SEED dataset such as PSD, DE, RASM, DASM, and ASM for the classification of emotions. LSTM model is then used for performing the classification based on these features.