Yoon Sungwook, Kim Byungmun
Gyeongbuk Development Institute, 201 Docheong-daero, Homyeong-eup, Yecheon 36849, Gyeongsangbuk-do, Republic of Korea.
Dept. of Electronic and Mechanical Engineering, GyeongKuk National University, 1375 Gyeongdong-ro, Andong 36729, Gyeongsangbuk-do, Republic of Korea.
Sensors (Basel). 2025 Aug 14;25(16):5066. doi: 10.3390/s25165066.
This paper introduces EmotionTFN (Emotion-Multi-Scale Temporal Fusion Network), a novel hierarchical temporal fusion architecture that addresses key challenges in IoT emotion recognition by processing diverse sensor data while maintaining accuracy across multiple temporal scales. The architecture integrates physiological signals (EEG, PPG, and GSR), visual, and audio data using hierarchical temporal attention across short-term (0.5-2 s), medium-term (2-10 s), and long-term (10-60 s) windows. Edge computing optimizations, including model compression, quantization, and adaptive sampling, enable deployment on resource-constrained devices. Extensive experiments on MELD, DEAP, and G-REx datasets demonstrate 94.2% accuracy on discrete emotion classification and 0.087 mean absolute error on dimensional prediction, outperforming the best baseline (87.4%). The system maintains sub-200 ms latency on IoT hardware while achieving a 40% improvement in energy efficiency. Real-world deployment validation over four weeks achieved 97.2% uptime and user satisfaction scores of 4.1/5.0 while ensuring privacy through local processing.
本文介绍了EmotionTFN（情感多尺度时间融合网络），这是一种新颖的分层时间融合架构，通过处理多样化的传感器数据，同时在多个时间尺度上保持准确性，解决了物联网情感识别中的关键挑战。该架构使用分层时间注意力，在短期（0.5 - 2秒）、中期（2 - 10秒）和长期（10 - 60秒）窗口内整合生理信号（脑电图、光电容积脉搏波描记图和皮肤电反应）、视觉和音频数据。包括模型压缩、量化和自适应采样在内的边缘计算优化，使得能够在资源受限的设备上进行部署。在MELD、DEAP和G - Rex数据集上进行的大量实验表明，在离散情感分类方面准确率达到94.2%，在维度预测方面平均绝对误差为0.087，优于最佳基线（87.4%）。该系统在物联网硬件上保持低于200毫秒的延迟，同时能源效率提高了40%。四周的实际部署验证实现了97.2%的正常运行时间和4.1/5.0的用户满意度评分，同时通过本地处理确保了隐私。
