Traffic Scene Similarity: a Graph-based Contrastive Learning Approach

TL;DR

This paper introduces a graph-based contrastive learning method to embed traffic scenes, enabling the identification of similar scenarios and reducing redundant testing in automated driving validation.

Contribution

It extends contrastive learning with graph representations to effectively cluster and analyze traffic scenes for scenario-based testing.

Findings

Scenes are continuously mapped in an embedding space

The approach forms thematically similar scene clusters

Redundant test scenarios can be reduced using clustering

Abstract

Ensuring validation for highly automated driving poses significant obstacles to the widespread adoption of highly automated vehicles. Scenario-based testing offers a potential solution by reducing the homologation effort required for these systems. However, a crucial prerequisite, yet unresolved, is the definition and reduction of the test space to a finite number of scenarios. To tackle this challenge, we propose an extension to a contrastive learning approach utilizing graphs to construct a meaningful embedding space. Our approach demonstrates the continuous mapping of scenes using scene-specific features and the formation of thematically similar clusters based on the resulting embeddings. Based on the found clusters, similar scenes could be identified in the subsequent test process, which can lead to a reduction in redundant test runs.