Fatigue classification method based on four-dimensional attention convolutional recurrent neural network

Abstract

The invention discloses a fatigue classification method based on a four-dimensional attention convolutional recurrent neural network, and the method comprises the following steps: S1, collecting an electroencephalogram signal, inputting the electroencephalogram signal into a four-dimensional feature extraction module, and extracting the four-dimensional features of the electroencephalogram signal; s2, inputting the extracted four-dimensional features into an attention module to obtain features with space-channel attention; and S3, inputting the features with space-channel attention into a convolutional recurrent neural network module, and performing fatigue classification. According to the method, the problem of poor interpretability of a neural network based on electroencephalogram signals is solved, the classification accuracy is improved, visualization can be carried out from the angles of space and frequency bands, and the interpretability of the network is improved. The depth separable convolution layer is used, compared with a common convolution layer, the size of the model is reduced by about 70%, the accuracy rate is improved by 1.44%, double-branch depth separable convolution is provided, receptive fields of two scales are fused on the aspect of spatial information processing, and the accuracy rate is further improved by 0.45%.

Description

technical field [0001] The invention relates to a fatigue classification method based on a four-dimensional attention convolutional neural network. Background technique [0002] Driving fatigue, usually caused by excessive activity and lack of rest, impairs the driver's ability to control the vehicle and has become one of the main causes of traffic accidents. According to statistics, about 20%-30% of traffic accidents are caused by fatigued driving, and about 60% of people admit to experiencing fatigued driving. According to the National Highway Traffic Safety Administration (NHTSA), every year is caused by fatigued driving. about 100,000 traffic accidents. Therefore, reliable fatigue detection has positive significance for traffic safety. [0003] According to previous studies, there are mainly three methods for monitoring fatigue driving. The first is a psychology-based method, which usually relies on psychometric questionnaires to assess an individual’s fatigue level; ...

AI Technical Summary

Problems solved by technology

The performance of 3-D CNN is promising in many applications, but the neural network still has problems such as insufficient input feature domain dimension, unreasonable model parameter amount and poor interpretability in EEG signal-based fatigue detection.

Images