Probabilistic Collision Risk Estimation for Pedestrian Navigation

2025-06-19

AI Summary

This study addresses the insufficient accuracy of collision warning for visually impaired pedestrians by pioneering the adaptation of probabilistic risk modeling—originally developed for autonomous driving—into wearable assistive devices. Methodologically, it fuses RGB-D sensing, inertial measurement unit (IMU) data, and real-time optical flow tracking to estimate dynamic object trajectories and compute time-varying collision probabilities, replacing conventional static distance or time-to-collision (TTC) metrics. The key contributions are: (1) the first lightweight, pedestrian-adapted probabilistic risk modeling framework derived from automotive safety paradigms; and (2) empirical validation on real-world data, achieving 67% hazardous-object detection accuracy—significantly outperforming distance-based (51%) and TTC-based (51%) baselines by 16 percentage points. Results demonstrate the superiority and practicality of probabilistic risk assessment in complex, dynamic urban environments for accessible navigation.

Technology Category

• Intelligent Robots: State Estimation
• Reasoning under Uncertainty: Relational Probabilistic Models
• Humans and AI: AI for Accessibility

Application Category

• Security and Privacy: Large-scale security measurements
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems

Abstract

Intelligent devices for supporting persons with vision impairment are becoming more widespread, but they are lacking behind the advancements in intelligent driver assistant system. To make a first step forward, this work discusses the integration of the risk model technology, previously used in autonomous driving and advanced driver assistance systems, into an assistance device for persons with vision impairment. The risk model computes a probabilistic collision risk given object trajectories which has previously been shown to give better indications of an object's collision potential compared to distance or time-to-contact measures in vehicle scenarios. In this work, we show that the risk model is also superior in warning persons with vision impairment about dangerous objects. Our experiments demonstrate that the warning accuracy of the risk model is 67% while both distance and time-to-contact measures reach only 51% accuracy for real-world data.

Problem

• Estimating collision risk for visually impaired pedestrians
• Improving warning accuracy for dangerous objects
• Integrating risk model technology into assistance devices

Innovation

• Integrates risk model from autonomous driving
• Estimates probabilistic collision risk accurately
• Superior warning accuracy for vision impairment