Machine learning for social coordination of autonomous vehicles with human road users

作为交通研究所(ITS)对可持续交通系统和改善行人安全的研究活动的一部分，这项研究将侧重于了解行人和脆弱的人类道路使用者与自动车辆(AVs)的社会互动，以提高城市交通的安全性，并降低由于AVs的错误决策和操纵而导致的事故率。利用公开的数据集和利兹大学的驾驶模拟器数据，该研究试图理解行人在交通场景中的行为，提高现有基于机器视觉的模型的准确性，并有助于填补该领域的学术研究空白，从而提高自动驾驶功能的性能和安全性。该学生还可以与其他具有心理学和人为因素背景的ITS研究人员合作，以更好地理解车辆和行人之间的相互作用。我们在利兹的人为因素和安全研究小组的所有研究活动都旨在实现一个共同的目标：保持弹性和更安全的交通系统，同时减少交通事故、死亡和受伤的数量。

As part of the research activities in the Institute for Transport Studies (ITS) toward sustainable transportation systems and improving the safety of pedestrians and, the research will focus on understanding the social interaction of pedestrians and vulnerable human road users with automated vehicles (AVs) to improve the safety of urban traffic and the lower the rate of accidents due to false decision making and manoeuvre by the AVs. Using publicly available datasets and the University of Leeds driving simulator data, the research tries to understand pedestrians' behaviour in traffic scenes, enhance the accuracy of the existing machine vision-based models and contributes to the academic research gaps in the field; hence improving the performance and safety of automated driving functions. The student may also collaborate with other researchers in ITS with a background in psychology and human factor for a better understanding of vehicle/pedestrian interactions. All research activities in our human factors and safety research group at ITS Leeds aim at one common objective: maintaining resilient and safer transportation systems while reducing the number of traffic accidents, fatalities, and injuries.