In order to activate the scientific research atmosphere of the college and promote academic exchanges between teachers and students, the College of Automotive organized the "Weekly Lecture" series of activities. This activity was held on May 31, 2023 at 15:00 in the conference room 311 of the School of Automotive, Ms. Ge Huimin from the Department of Transportation and Communications of the School of Automotive delivered an academic lecture on the theme of "Advances in Driving Distraction Research", which was presided over by Mr. Wang Ruochen, Vice Dean of the School of Automotive.

Taking driving behavior as the main research object and combining statist

ical methods, the acquisition of driving distraction test data is systematically described. Experiment is the main way of driving distraction data acquisition, natural driving datasets and video recordings are new data acquisition methods, and roadside observation and questionnaires are less concerned; the current types and topics of driving distraction research are more concentrated, fusion indicators are the most frequently used driving distraction identification indicators, and the two types of fusion indicators, namely, driving performance indicators and eye-movement indicators, and driving performance indicators and reaction indicators, are more frequently used. Driving performance indicator is the most used single class indicator; support vector machine model is currently the most used driving distraction recognition model, but the recognition accuracy standard deviation is large and the performance is unstable, deep learning algorithm model has higher recognition accuracy and good stability. Future driving distraction research needs to balance the research topics, expand the distraction scenarios of human-machine co-driving, deepen the research on driving distraction types, and construct a standardized indicator system and selection principles.