Affiliation:
1. University of Michigan
Abstract
Distracted driving is a leading cause of accidents worldwide. The tasks of distraction detection and recognition have been traditionally addressed as computer vision problems. However, distracted behaviors are not always expressed in a visually observable way. In this work, we introduce a novel multimodal dataset of distracted driver behaviors, consisting of data collected using twelve information channels coming from visual, acoustic, near-infrared, thermal, physiological and linguistic modalities. The data were collected from 45 subjects while being exposed to four different distractions (three cognitive and one physical). For the purposes of this paper, we performed experiments with visual, physiological, and thermal information to explore potential of multimodal modeling for distraction recognition. In addition, we analyze the value of different modalities by identifying specific visual, physiological, and thermal groups of features that contribute the most to distraction characterization. Our results highlight the advantage of multimodal representations and reveal valuable insights for the role played by the three modalities on identifying different types of driving distractions.
Funder
Toyota Research Institute
Publisher
Association for Computing Machinery (ACM)
Subject
Artificial Intelligence,Human-Computer Interaction
Reference55 articles.
1. Detecting Deceptive Behavior via Integration of Discriminative Features From Multiple Modalities
2. Real-time distracted driver posture classification;Abouelnaga Yehya;arXiv preprint arXiv:1706.09498,2017
3. W. Avinash, S. Dvijesh, and P. Ioannis. 2010. A novel method to monitor driver’s distractions. In Proceedings of the 28th International Conference Extended Abstracts on Human Factors in Computing Systems, Atlanta, Georgia, USA.
4. OpenFace: An open source facial behavior analysis toolkit
5. Intelligent driver monitoring systems based on physiological sensor signals: A review
Cited by
4 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献