The Influence Of Traffic Semantic Information Types On Drivers’ Visual Attention

Individuals can improve their visual search efficiency by acquiring stable associations between objects in a given environment,an effect collectively known as“context effect”.In real-world driving situations,key traffic information is usually not isolated but always interconnected with other information,and the recognition of an object is always closely related to the information in its environment.The present study investigates the influence of semantic relationships between traffic information on drivers’ visual attention in dynamic scenarios by using two eye-tracking experiments with young drivers as the subjects.In the first experiment,20 subjects of different genders were selected to perform an unintentional blindness task,in which they were asked to imagine themselves as traffic police officers,watch the real-life driving videos presented in the experiment from the perspective of a traffic police officer and answer the questions that appeared after each video.Each participant was asked to watch a total of 20 videos,in the order of 10 videos in the first part,and then continue watching 10 videos in the second part after a break.Of the 10 videos in each part,only the last video content and the post-video question were related to the dependent attentional cue(cued/uncued)and eye movement data and verbal responses were recorded.The results showed that drivers’ total gaze time and first gaze time on the target(target traffic sign)were longer in the video with attentional cues than in the video without cues;female drivers’ total gaze time on the target was longer in the video with attentional cues than in the video without cues.In the second experiment,19 subjects of different genders were selected to perform a driving unintentional blindness task with the same tasks and requirements as in Experiment 1.Of the 10 videos in each section,only the content of the last video and the post-video question were related to the semantic relationship(semantic agreement/semantic inconsistency)of the independent variable and eye movement data and subjects’ verbal responses were recorded.The results showed that male drivers showed a significant negative correlation between the total gaze time of targets(target traffic signs)and distractors(groups of distracting traffic signs)in the semantic incongruent videos,but did not show a significant correlation between the total gaze time of targets and distractors in the semantic congruent videos.According to the results of this study,the following conclusions are derived:(1)different semantic relationships between traffic information have different effects on drivers’ driving attention,however,this effect is based on the premise that drivers identify the semantic information in a relationship as relevant to the task at hand;(2)female drivers are more cautious when processing driving information and tend to invest more time in obtaining relevant driving information.Male drivers are more efficient at extracting traffic information and are better at adapting driving attention through semantic relationships between traffic information.