Close

Presentation

60. Uncertainty Differences in Computing Hierarchical Pedestrian Behaviors
DescriptionThe emergence of autonomous driving technology promises a paradigm shift in urban dynamics, yet predicting pedestrian behavior remains a critical challenge. This study investigates the complexities of pedestrian behavior, focusing on two distinct types of uncertainties: aleatoric and epistemic, aiming to dissect their distribution and correlation across different pedestrian behaviors. Leveraging an Evidential Deep Learning algorithm with high training performance on the PIE pedestrian behavior benchmark dataset, the research examines pedestrian road crossing and short-term movement selection behaviors. Findings reveal notable differences in uncertainty distribution. Aleatoric uncertainty, reflecting inherent pedestrian behavior randomness, is consistently lower than epistemic uncertainty, which also includes perception limitations. This suggests that pedestrians themselves exhibit less uncertainty in their actions compared to the uncertainty faced by external observers. Moreover, the study demonstrates that road-crossing predictions exhibit significantly larger uncertainties compared to short-term destination predictions, highlighting the challenges in anticipating more complex pedestrian behaviors.
Event Type
Poster
TimeThursday, September 12th5:30pm - 6:30pm MST
LocationMcArthur Ballroom
Tracks
Aging
Augmented Cognition
Children's Issues
Communications
Cybersecurity
Education
Environmental Design
General Sessions
Human AI Robot Teaming (AI)
Macroergonomics
Occupational Ergonomics
Student Forum
Surface Transportation
Sustainability
System Development