Human data collection protocol for image-based driver monitoring system validation

用于基于图像的驾驶员监控系统验证的人体数据收集协议

• 批准号: 506315-2016
• 负责人: Cao, Shi
• 金额: $ 1.82万
• 依托单位: University of Waterloo
• 依托单位国家: 加拿大
• 项目类别: Engage Grants Program
• 财政年份: 2016
• 资助国家: 加拿大
• 起止时间: 2016-01-01 至 2017-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=610158
• 关键词: Human; data; collection; protocol; image

Computer vision technologies have been developed to monitor drivers' behaviours, detecting unsafe events such as fatigue and distraction. Before the technology is released to the public, its accuracy and robustness must be tested with human data collected from a wide range of driver populations and driver conditions. To build a large data base, an appropriate testing protocol is important. It should be scientifically sound and well designed for efficiency. This project aims to design and validate a human data collection protocol for this purpose. It will also address research questions related to driving safety. For example, how to determine the impact of critical driver behaviours on driving safety and design proper warnings?

计算机视觉技术已被开发用于监控驾驶员的行为，检测疲劳和分心等不安全事件。在向公众发布该技术之前，必须使用从广泛的驾驶员人群和驾驶员条件中收集的人类数据对其准确性和鲁棒性进行测试。为了建立一个大的数据库，一个合适的测试协议是很重要的。它应该是科学合理的，设计良好的效率。该项目旨在为此目的设计和验证人体数据收集协议。它还将解决与驾驶安全有关的研究问题。例如，如何确定关键驾驶员行为对驾驶安全的影响并设计适当的警告？