多模态驾驶员监控数据库(MDM)
收藏资源简介:
多模态驾驶员监控数据库(MDM)是由德州大学达拉斯分校电气与计算机工程系创建的一个自然语言语料库,旨在研究驾驶员的注意力。该数据集包含59名受试者在执行各种任务时的记录,使用多种传感器收集高质量的RGB视频、点云数据、CAN-Bus信息和音频记录,以训练各种高效的驾驶员监控算法,进一步推动车内安全系统领域的发展。数据集创建过程中,采用了多种设备和传感器来设置基线和记录数据,并通过详细的协议来获取标记数据,以研究驾驶员的视觉注意力。该数据集的应用领域主要集中在驾驶员行为建模和车内安全系统的开发,旨在解决驾驶员分心和人为错误导致的事故问题。
Multimodal Driver Monitoring Database (MDM) is a natural language corpus developed by the Department of Electrical and Computer Engineering, The University of Texas at Dallas, for the purpose of studying driver attention. This dataset includes recordings from 59 subjects while they performed various tasks, with high-quality RGB videos, point cloud data, CAN-Bus information and audio recordings collected via multiple sensors, to train a variety of efficient driver monitoring algorithms and further promote the development of in-vehicle safety systems. During the dataset creation process, multiple devices and sensors were employed to establish baselines and record data, and detailed protocols were utilized to acquire labeled data for research on driver visual attention. The application scenarios of this dataset mainly concentrate on driver behavior modeling and the development of in-vehicle safety systems, with the goal of addressing traffic accidents caused by driver distraction and human error.

- 1The Multimodal Driver Monitoring Database: A Naturalistic Corpus to Study Driver Attention德州大学达拉斯分校电气与计算机工程系 · 2020年



