FuzzRisk: Online Collision Risk Estimation for Autonomous Vehicles based on Depth-Aware Object Detection via Fuzzy Inference

TL;DR

This paper introduces FuzzRisk, a fuzzy inference-based framework that estimates collision risk for autonomous vehicles by analyzing inconsistencies between depth-aware and standard object detection predictions, validated on large-scale datasets.

Contribution

The paper proposes a novel online collision risk estimation method combining depth-aware and standard object detection with fuzzy inference, improving AV safety monitoring.

Findings

Inconsistencies between detection methods correlate with detector errors.

FuzzRisk's risk estimates align with actual collision rates.

Framework effectively safeguards AVs in simulations.

Abstract

This paper presents a novel monitoring framework that infers the level of collision risk for autonomous vehicles (AVs) based on their object detection performance. The framework takes two sets of predictions from different algorithms and associates their inconsistencies with the collision risk via fuzzy inference. The first set of predictions is obtained by retrieving safety-critical 2.5D objects from a depth map, and the second set comes from the ordinary AV's 3D object detector. We experimentally validate that, based on Intersection-over-Union (IoU) and a depth discrepancy measure, the inconsistencies between the two sets of predictions strongly correlate to the error of the 3D object detector against ground truths. This correlation allows us to construct a fuzzy inference system and map the inconsistency measures to an AV collision risk indicator. In particular, we optimize the fuzzy inference system towards an existing offline metric that matches AV collision rates well. Lastly, we validate our monitor's capability to produce relevant risk estimates with the large-scale nuScenes dataset and demonstrate that it can safeguard an AV in closed-loop simulations.