Traffic accident-related injuries and fatalities are a serious global public health and social development challenge. The accurate prediction of crash severity improves road safety and reduces casualties, as well as serving road managers and policy makers. Prediction models need to learn and analyze the various characteristic factors of traffic accidents and capture from them the inherent complex relationship between accident characteristics and the severity of traffic accidents. However, most accident prediction studies lack analytical predictions of injury severity, and predictive models rely on the content and quality of accident datasets. To increase the robustness and accuracy of prediction models, this paper leverages a Transformer-based architecture for the severity prediction of traffic collisions from human injury severity. This framework learns both text and sequence data from accident datasets. After comparative analysis, the framework can achieve the prediction of human injury severity under different data categories and show good prediction performance at low injury severity levels using only textual data or sequence data.
23 July 2025 in Signal, Image and Video Processing
With the popularization of social media andInternet technology, people are more and more easilydistracted when driving, which has a serious adverseimpact on the safety of expected functions, and reducesthe efficiency of control transmission between humanand computer. This article proposes a real-time pupil re-construction (PURE) algorithm for real-time detectionof pupil direction, and combines it with a 3D eye modelto obtain the gaze direction of the pupil in real time.Subsequently, the Posit algorithm was used to obtain real-time head posture, and the head posture and pupil gaze direction were used as improved input features toestimate the driver’s gaze area in real-time using a random forest model. Finally, this article comprehensivelyconsiders the impact of multiple real-time indicatorson driving attention, such as the number of eye gazeareas, eye scanning areas, road conditions, brain latency,and driving tasks, and proposes a prototype framework for real-time driver multi indicator attention evaluation.Through various real-time evaluation experiments,we have demonstrated that: 1) these context relatedattention features are universal when tested in real-timeon 16 drivers in a driving simulator; 2) The proposeddriver attention evaluation model performs well underreal-time conditions and can produce reliable results invarious driving scenarios.