Automatic driver stress level classification using multimodal deep learning

作者：

Highlights：

• Multimodal fusion model based on CNN-LSTM network to recognize driver stress level.

• First deep learning approach applied to ECG, vehicle data and environmental data.

• Multimodal deep learning approach is effective in detecting driver stress level.

• Fusion approach using CNN-LSTM performs better than handcrafted feature extraction.

摘要

•Multimodal fusion model based on CNN-LSTM network to recognize driver stress level.•First deep learning approach applied to ECG, vehicle data and environmental data.•Multimodal deep learning approach is effective in detecting driver stress level.•Fusion approach using CNN-LSTM performs better than handcrafted feature extraction.

论文关键词：Deep learning,Driver stress detection,Convolutional neural network,Long short term memory,ECG signal,Vehicle data

论文评审过程：Received 26 February 2019, Revised 30 June 2019, Accepted 5 July 2019, Available online 6 July 2019, Version of Record 25 July 2019.

论文官网地址：https://doi.org/10.1016/j.eswa.2019.07.010