STAFNet: an adaptive multi-feature learning network via spatiotemporal fusion for EEG-based emotion recognition DOI Creative Commons
Fo Hu, Kailun He,

Mengyuan Qian

et al.

Frontiers in Neuroscience, Journal Year: 2024, Volume and Issue: 18

Published: Dec. 10, 2024

Introduction Emotion recognition using electroencephalography (EEG) is a key aspect of brain-computer interface research. Achieving precision requires effectively extracting and integrating both spatial temporal features. However, many studies focus on single dimension, neglecting the interplay complementarity multi-feature information, importance fully dynamics to enhance performance. Methods We propose Spatiotemporal Adaptive Fusion Network (STAFNet), novel framework combining adaptive graph convolution transformers accuracy robustness EEG-based emotion recognition. The model includes an convolutional module capture brain connectivity patterns through dynamic evolution multi-structured transformer fusion integrate latent correlations between features for classification. Results Extensive experiments were conducted SEED SEED-IV datasets evaluate performance STAFNet. achieved accuracies 97.89% 93.64%, respectively, outperforming state-of-the-art methods. Interpretability analyses, including confusion matrices t-SNE visualizations, employed examine influence different emotions model's Furthermore, investigation varying GCN layer depths demonstrated that STAFNet mitigates over-smoothing issue in deeper architectures. Discussion In summary, findings validate effectiveness results emphasize critical role spatiotemporal feature extraction introduce innovative fusion, advancing state art

Language: Английский

A discriminative multi-modal adaptation neural network model for video action recognition DOI
Lei Gao, Kai Liu, Ling Guan

et al.

Neural Networks, Journal Year: 2025, Volume and Issue: 185, P. 107114 - 107114

Published: Jan. 5, 2025

Language: Английский

Citations

1
Emotion Recognition in Human-Machine Interaction and a Review in Interpersonal Communication Perspective DOI
Vimala Govindaraju, Dhanabalan Thangam

Advances in computational intelligence and robotics book series, Journal Year: 2024, Volume and Issue: unknown, P. 329 - 343

Published: June 30, 2024

Emotions are fundamental to daily decision-making and overall wellbeing. psychophysiological processes that frequently linked human-machine interaction, it is expected we will see the creation of systems can recognize interpret human emotions in a range ways as computers computer-based applications get more advanced pervasive people's lives. Emotion recognition able modify their responses user experience based on analysis interpersonal communication signals. The ability virtual assistants respond emotionally effectively, support mental health by identifying users' emotional states, enhancement interaction applications. aim this chapter reviewing elements models now.

Language: Английский

Citations

4
Comprehensive Survey on Recognition of Emotions from Body Gestures DOI Open Access

Ramakrishna Gandi

Journal of Informatics Education and Research, Journal Year: 2025, Volume and Issue: 5(1)

Published: Jan. 17, 2025

Automatic emotion identification has emerged as a prominent area of research during the past decade, with applications in healthcare, human-computer interaction, and behavioral analysis. Although facial expressions verbal communication have been thoroughly examined, emotions via body gestures is still inadequately investigated. Body gestures, an essential aspect "body language" offer significant contextual indicators shaped by gender culture variations. Recent breakthroughs deep learning facilitated development robust models capable accurately capturing complex human movements, hence enhancing recognition precision adaptability. This study presents thorough framework for automatic emotional encompassing elements such individual detection, position estimation, representation learning. High computational costs need advanced algorithms to fuse multimodal data add these hurdles. advancements learning, shown great potential overcome issues improve accuracy. work highlights applications, challenges, future directions from emphasizing scalable, robust, real-world-ready systems that can enable emotionally intelligent technologies.

Language: Английский

Citations

0
MSF-Net: Multi-stage fusion network for emotion recognition from multimodal signals in scalable healthcare DOI Creative Commons
Md. Milon Islam,

Fakhri Karray,

Ghulam Muhammad

et al.

Information Fusion, Journal Year: 2025, Volume and Issue: unknown, P. 103028 - 103028

Published: Feb. 1, 2025

Language: Английский

Citations

0
Multimodal Emotion Recognition based on Face and Speech using Deep Convolution Neural Network and Long Short Term Memory DOI
Shwetkranti Taware, Anuradha Thakare

Circuits Systems and Signal Processing, Journal Year: 2025, Volume and Issue: unknown

Published: April 25, 2025

Language: Английский

Citations

0
Generalized multisensor wearable signal fusion for emotion recognition from noisy and incomplete data DOI

Vamsi Kumar Naidu Pallapothula,

Sidharth Anand, Sreyasee Das Bhattacharjee

et al.

Smart Health, Journal Year: 2025, Volume and Issue: unknown, P. 100571 - 100571

Published: March 1, 2025

Language: Английский

Citations

0
Leveraging Emotional AI for Improved Human-Computer Interactions DOI
Vimala Govindaraju, Dhanabalan Thangam

Advances in computational intelligence and robotics book series, Journal Year: 2024, Volume and Issue: unknown, P. 66 - 81

Published: June 6, 2024

Emotions are psychophysiological processes that sparked by both conscious and unconscious perceptions of things events. Mood, motivation, temperament, personality frequently linked to emotions. Human-machine interaction will see the creation systems can recognize interpret human emotions in a range ways as computers computer-based applications get more advanced pervasive people's daily lives. More sympathetic customized relationships between humans machines result from efficient emotion recognition human-machine interactions. Emotion able modify their responses user experience based on analysis interpersonal communication signals. The ability virtual assistants respond emotionally effectively, support mental health identifying users' emotional states, improvement customer interactions with responsive Chabots, enhancement human-robot collaboration just few examples real-world applications. Reviewing elements models now use is aim this chapter.

Language: Английский

Citations

3
A Model of Sentiment Analysis for College Music Teaching Based on Musical Expression DOI Creative Commons
Xuecheng Wang

Applied Mathematics and Nonlinear Sciences, Journal Year: 2024, Volume and Issue: 9(1)

Published: Jan. 1, 2024

Abstract In this paper, we first present the structure of Hierarchical Sentiment Analysis Model for Multimodal Fusion (HMAMF). The model uses Bi-LSTM method to extract unimodal music features and a CME encoder feature fusion. After sentiment analysis, loss function auxiliary training dataset is obtained co-trained. Finally, application HMAMF in university teaching being explored. results show that agreement between dominant prediction >80%, well-tested. underwent 35 sessions when correct rate network recognition was 97.19%. mean accuracy model’s 3-time lengths from 50 seconds 300 ranged 87.92% 98.20%, there slight decrease as length increased. mood beat were judged by way highly consistent with students’ delineation results. Students teachers’ satisfaction performance analysis terms “music tempo, rhythm, mood, content, time” 81.15% 85.83% 83.25% 92.39%, respectively. Teachers students are satisfied proposed paper at 89.43% 90.97%, proven be suitable use process.

Language: Английский

Citations

0
Facial Emotion Recognition for Enhanced Human-Computer Interaction using Deep Learning and Temporal Modeling with BiLSTM DOI
Parasuraman Karthikeyan,

S Kirutheesvar,

S. Sivakumar

et al.

Published: Sept. 18, 2024

Language: Английский

Citations

0
STAFNet: an adaptive multi-feature learning network via spatiotemporal fusion for EEG-based emotion recognition DOI Creative Commons
Fo Hu, Kailun He,

Mengyuan Qian

et al.

Frontiers in Neuroscience, Journal Year: 2024, Volume and Issue: 18

Published: Dec. 10, 2024

Introduction Emotion recognition using electroencephalography (EEG) is a key aspect of brain-computer interface research. Achieving precision requires effectively extracting and integrating both spatial temporal features. However, many studies focus on single dimension, neglecting the interplay complementarity multi-feature information, importance fully dynamics to enhance performance. Methods We propose Spatiotemporal Adaptive Fusion Network (STAFNet), novel framework combining adaptive graph convolution transformers accuracy robustness EEG-based emotion recognition. The model includes an convolutional module capture brain connectivity patterns through dynamic evolution multi-structured transformer fusion integrate latent correlations between features for classification. Results Extensive experiments were conducted SEED SEED-IV datasets evaluate performance STAFNet. achieved accuracies 97.89% 93.64%, respectively, outperforming state-of-the-art methods. Interpretability analyses, including confusion matrices t-SNE visualizations, employed examine influence different emotions model's Furthermore, investigation varying GCN layer depths demonstrated that STAFNet mitigates over-smoothing issue in deeper architectures. Discussion In summary, findings validate effectiveness results emphasize critical role spatiotemporal feature extraction introduce innovative fusion, advancing state art

Language: Английский

Citations

0