Driver’s mobile phone usage detection using guided learning based on attention features and prior knowledge

作者：

Highlights：

• DMPU is detected using guided learning and prior knowledge.

• AAM/3DW is used to calculate the importance of each neuron.

• Guided learning refers to the use of attention features as constraints.

• The prior knowledge is used as part of the weight.

• Feature visualization method is proposed to visualize features related to DMPU.

摘要

•DMPU is detected using guided learning and prior knowledge.•AAM/3DW is used to calculate the importance of each neuron.•Guided learning refers to the use of attention features as constraints.•The prior knowledge is used as part of the weight.•Feature visualization method is proposed to visualize features related to DMPU.

论文关键词：Mobile phone usage,Attention mechanism,Attention feature,Prior knowledge,Deep neural network

论文评审过程：Received 26 December 2021, Revised 1 April 2022, Accepted 13 June 2022, Available online 16 June 2022, Version of Record 18 June 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.117877