Synthetic Distracted Driving (SynDD2) dataset
收藏资源简介:
Synthetic Distracted Driving (SynDD2)数据集由爱荷华州立大学创建,旨在通过机器学习模型分析驾驶员的分心行为和不同注视区域。该数据集包含两种活动类型:分心活动和注视区域,每种活动类型又分为有无外观遮挡两组。数据集通过三台车内摄像头采集,包含视频文件和注释文件,用于评估机器学习算法在分类驾驶员分心活动和注视区域方面的性能。创建过程中,参与者在静止车辆中执行随机顺序和持续时间的活动。该数据集主要应用于开发驾驶员辅助系统,以提高驾驶安全性。
Synthetic Distracted Driving (SynDD2) dataset was developed by Iowa State University, with the goal of analyzing driver distraction behaviors and different gaze regions using machine learning models. This dataset covers two categories of activities: distraction-related activities and gaze regions, each of which is further divided into two subgroups with and without visual occlusion. Collected via three in-vehicle cameras, the dataset includes video files and annotation files, which are used to evaluate the performance of machine learning algorithms in classifying driver distraction activities and gaze regions. During the dataset's creation, participants performed activities with randomized order and duration in a stationary vehicle. This dataset is primarily applied to develop driver assistance systems to enhance driving safety.




