On-Board Pedestrian and Cyclist Recognition Using Deep Learning Methods

In recent years, the application of advanced driver assistance systems (ADAS) has improved road safety. However, unlike motor vehicle drivers, pedestrians, cyclists, and other vulnerable road users have been given less attention. Most existing pedestrian and cyclist recognition methods recognize them separately, which often leads to confusion between the two classes. Also, the limited capability of traditional object recognition models makes them difficult to apply in complicated road environments. This paper establishes a unified framework for pedestrian and cyclist concurrent recognition, with relevant key technologies in the recognizing process, including multiple-instance object proposal method, deep neural network-based object detection, and multiple object tracking with online target-specific self-learning function. Experimental results indicate that the proposed pedestrian and cyclist concurrent recognition method can not only detect pedestrians and cyclists effectively, but also differentiate them clearly, which will provide new information for decision-making in intelligent vehicles.

• Record URL:
  • https://doi.org/10.1061/9780784481523.016
• Availability:
  • Find a library where document is available. Order URL: http://worldcat.org/isbn/9780784481523
• Supplemental Notes:
  • © 2018 American Society of Civil Engineers.
• Corporate Authors:

  American Society of Civil Engineers

  1801 Alexander Bell Drive
  Reston, VA  United States  20191-4400
• Authors:
  • Chen, Wenqiang
  • Xiong, Hui
  • Li, Keqiang
  • Li, Xiaofei
  • Zhang, Dezhao
• Conference:
  • 18th COTA International Conference of Transportation Professionals
  • Location: Beijing , China
  • Date: 2018-7-5 to 2018-7-8
• Publication Date: 2018-7

Language

• English

Media Info

• Media Type: Web
• Features: References;
• Pagination: pp 169-175
• Monograph Title: CICTP 2018: Intelligence, Connectivity, and Mobility

Subject/Index Terms

• TRT Terms: Cyclists; Detection and identification systems; Machine learning; Neural networks; Pedestrians
• Subject Areas: Data and Information Technology; Highways; Pedestrians and Bicyclists; Vehicles and Equipment;

Filing Info

• Accession Number: 01870398
• Record Type: Publication
• ISBN: 9780784481523
• Files: TRIS, ASCE
• Created Date: Jan 23 2023 12:22PM