Scholarlink

<p>Emotions have a significant effect on daily living and they are linked to physical as well as mental wellness. People use words, noises, body language and facial expressions to convey their feelings. Real feelings, however, are sometimes difficult to explain through such behaviors since people can occasionally alter the emotions that are exhibited. Emotional analysis and comprehension are so crucial. Electroencephalography (EEG signals) can be used to identify emotional states.EEG-based emotion recognition provides vital insights into emotional states. In today’s world, when mental health interventions and adaptive technologies are becoming more and more prevalent due to the rising incidence of mental health issues and the need for individualized experiences. This work is aimportant step forward in linking physiological data and real-time emotional comprehension, which is essential for improving user experiences and therapeutic results. In this article, we present a Deep Learning (DL) method to identify the emotions in EEG data and acquire their attributes concurrently. In this research, we utilize a dataset gathered from numerous people who examined a series of visuals across several days through an EEG signal.Using a median filter approach, the data were filtered to decrease noise before feature extraction. To extract significant characteristics from preprocessed EEG data, Principle Component Analysis (PCA) is used. PCA reduces the dimensionality of the data while keeping its key qualities, allowing the extraction of distinctive characteristics connected with different emotional states to be facilitated. For classification approach, we use a novel deep learning method called Satin Bowerbird Optimization-enhanced Deep Recurrent Neural Network (SBBO-DRNN) is an effective classification mechanism which is utilized for emotion identification system. SBBO-DRNN is proposed for recognizing emotions from physiological data in EEG signals. The DRNN’s hyperparameters are optimized using the SBBO method, which improves the DRNN's ability to detect subtle patterns and temporal relationships in the EEG signals that correlate to various emotions. Several techniques are used to compare the suggested deep learning model. The experimental results showed Accuracy (97.12%), Precision (97.26%), Recall (96.63%) and F1-score (96.9%). The best effect detection accuracy is found in EEG characteristics from all physiological channels, according to the results. By the end of the study, our suggested approach outperforms other approaches in terms of accuracy.</p>

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

Recognizing Emotions from Physiological Data in a Eeg Signals Using a Novel Deep Learning Technique

  • 期刊论文

  • 发表日期：

    2025年1月25日

查看全文
  • R. Nandakumar,
  • S. Deivanayagi,
  • S. P. Angeline Kirubha,
  • R. Prabu

摘要

Emotions have a significant effect on daily living and they are linked to physical as well as mental wellness. People use words, noises, body language and facial expressions to convey their feelings. Real feelings, however, are sometimes difficult to explain through such behaviors since people can occasionally alter the emotions that are exhibited. Emotional analysis and comprehension are so crucial. Electroencephalography (EEG signals) can be used to identify emotional states.EEG-based emotion recognition provides vital insights into emotional states. In today’s world, when mental health interventions and adaptive technologies are becoming more and more prevalent due to the rising incidence of mental health issues and the need for individualized experiences. This work is aimportant step forward in linking physiological data and real-time emotional comprehension, which is essential for improving user experiences and therapeutic results. In this article, we present a Deep Learning (DL) method to identify the emotions in EEG data and acquire their attributes concurrently. In this research, we utilize a dataset gathered from numerous people who examined a series of visuals across several days through an EEG signal.Using a median filter approach, the data were filtered to decrease noise before feature extraction. To extract significant characteristics from preprocessed EEG data, Principle Component Analysis (PCA) is used. PCA reduces the dimensionality of the data while keeping its key qualities, allowing the extraction of distinctive characteristics connected with different emotional states to be facilitated. For classification approach, we use a novel deep learning method called Satin Bowerbird Optimization-enhanced Deep Recurrent Neural Network (SBBO-DRNN) is an effective classification mechanism which is utilized for emotion identification system. SBBO-DRNN is proposed for recognizing emotions from physiological data in EEG signals. The DRNN’s hyperparameters are optimized using the SBBO method, which improves the DRNN's ability to detect subtle patterns and temporal relationships in the EEG signals that correlate to various emotions. Several techniques are used to compare the suggested deep learning model. The experimental results showed Accuracy (97.12%), Precision (97.26%), Recall (96.63%) and F1-score (96.9%). The best effect detection accuracy is found in EEG characteristics from all physiological channels, according to the results. By the end of the study, our suggested approach outperforms other approaches in terms of accuracy.