GraphRQI: Classifying Driver Behaviors Using Graph Spectrums

TL;DR

GraphRQI is a new spectral graph-based algorithm that classifies driver behaviors from traffic data with improved accuracy and efficiency, aiding in understanding and predicting driver actions in urban environments.

Contribution

The paper introduces GraphRQI, a spectral analysis-based method for driver behavior classification that includes a novel eigenvalue algorithm with theoretical efficiency guarantees.

Findings

Achieves up to 25% higher accuracy than previous methods.

Runs twice as fast as existing eigenvalue algorithms.

Effectively predicts future trajectories of road-agents.

Abstract

We present a novel algorithm (GraphRQI) to identify driver behaviors from road-agent trajectories. Our approach assumes that the road-agents exhibit a range of driving traits, such as aggressive or conservative driving. Moreover, these traits affect the trajectories of nearby road-agents as well as the interactions between road-agents. We represent these inter-agent interactions using unweighted and undirected traffic graphs. Our algorithm classifies the driver behavior using a supervised learning algorithm by reducing the computation to the spectral analysis of the traffic graph. Moreover, we present a novel eigenvalue algorithm to compute the spectrum efficiently. We provide theoretical guarantees for the running time complexity of our eigenvalue algorithm and show that it is faster than previous methods by 2 times. We evaluate the classification accuracy of our approach on traffic videos and autonomous driving datasets corresponding to urban traffic. In practice, GraphRQI achieves an accuracy improvement of up to 25% over prior driver behavior classification algorithms. We also use our classification algorithm to predict the future trajectories of road-agents.