基于生理信号的情绪识别腕戴设备

Emotion recognition wristband device based on physiological signals

针对情感计算需求,设计了一种基于STM32L0的低功耗生理信号采集腕带设备,利用低功耗蓝牙无线通信将采集的生理信号实时发送至具有蓝牙4.0接口的智能设备端,采用BP神经网络对生理信号进行分析处理。实验结果表明,该设备可实现准确的心率、皮肤温度、皮肤阻抗、运动状态检测,通过多维度的生理信号分析,识别个体的情绪状态,其中紧张、中性、兴奋的识别率达到95%以上,为情感计算提供一种可穿戴设备。

For the requirement on affective computing, a STM32L0-based low power consumption wristband device to gather physiological signals is designed. Those collected physiological signals will be sent to smart device with Bluetooth 4.0 port through low power consumption Bluetooth wireless communication and analyzed by BP neural networks. Experiment result shows that this device can accurately detect heart rate, skin temperature, skin impedance and state of motion. This is a wearable device for affective computing which can identify emotional state of individuals with recognition rate of over 95% in tension, neutral and excitement through multidimensional physiological signal analysis.