Arousal and executive: multimodal physiological network for operator vigilance state detection
Description
data/train_data is used for model training. data/test_data for model testing code is the entire program for the training and testing process
Files
Steps to reproduce
EEG signals were collected using the Neuracle’s Neusen W-series wireless acquisition system, which supports synchronous wireless data acquisition at a sampling rate of 1000 Hz. According to the international 10-20 system, channels F3, F4, T3, T4, P3, P4, O1, and O2 were selected as the sampling electrodes, with physiological saline used as the conductive medium. ECG and EDA signals were collected using the ErgoLAB intelligent wearable human factors system, which consists of a comprehensive array of wearable sensors. The data sampling rate of this system is 64 Hz. Specifically, it includes an ear-clip sensor, a wrist sensor, a finger sensor, a chest strap sensor, and the ErgoLAB 3.0 signal processing system, which is connected to the computer via the ErgoLAB dongle. The experimental tasks utilized Steam VR, an HTC VIVE head-mounted VR device, and a multifunctional neuropsychological behavioral test trainer. Before the formal experiment, all subjects underwent a training session to familiarize themselves with the experimental environment, clarify the task procedures, practice each experimental task, and complete a proficiency test. EEG signals were downsampled to 64 Hz to standardize the sampling frequency across multimodal physiological signals. The data under each label for each operator were segmented according to different segment lengths, generating datasets of varying sizes. Specifically, signal segments of 1, 1.5, 2, 2.5, 3, 3.5, 4, 4.5, and 5 seconds were selected as samples for subsequent analysis. A total of ten channels were recorded, including eight EEG channels (F3, F4, T3, T4, P3, P4, O1, and O2), one ECG channel, and one EDA channel.
Institutions
- Taiyuan University of Technology