Correcting Autonomous Driving Object Detection Misclassifications with Automated Commonsense Reasoning

TL;DR

This work tackles misclassifications in autonomous vehicle perception by layering automated commonsense reasoning atop a deep learning perception stack. The authors introduce a five-step methodology that converts perception outputs into a knowledge base, then uses a logic-based layer to detect inconsistencies and correct classifications, enhancing safety and explainability. They implement BEV semantic segmentation, traffic-light detection, and behavior analysis, coupled with an evidential deep learning framework to quantify uncertainty, and integrate these with a Prolog-like commonsense module to handle scenarios such as malfunctioning signals and unseen obstacles. Experimental validation in the CARLA simulator demonstrates substantial performance gains over baseline perception models, both with a fully active reasoning layer and with uncertainty-triggered invocation, suggesting a viable path toward more reliable and explainable AV perception. The results indicate that hybrid perception systems combining DL with structured commonsense reasoning can improve safety-critical decision-making and provide a scalable route toward higher levels of autonomous driving autonomy.

Abstract

Autonomous Vehicle (AV) technology has been heavily researched and sought after, yet there are no SAE Level 5 AVs available today in the marketplace. We contend that over-reliance on machine learning technology is the main reason. Use of automated commonsense reasoning technology, we believe, can help achieve SAE Level 5 autonomy. In this paper, we show how automated common-sense reasoning technology can be deployed in situations where there are not enough data samples available to train a deep learning-based AV model that can handle certain abnormal road scenarios. Specifically, we consider two situations where (i) a traffic signal is malfunctioning at an intersection and (ii) all the cars ahead are slowing down and steering away due to an unexpected obstruction (e.g., animals on the road). We show that in such situations, our commonsense reasoning-based solution accurately detects traffic light colors and obstacles not correctly captured by the AV's perception model. We also provide a pathway for efficiently invoking commonsense reasoning by measuring uncertainty in the computer vision model and using commonsense reasoning to handle uncertain scenarios. We describe our experiments conducted using the CARLA simulator and the results obtained. The main contribution of our research is to show that automated commonsense reasoning effectively corrects AV-based object detection misclassifications and that hybrid models provide an effective pathway to improving AV perception.

Figures (2)

• Figure 1: General Procedure Architecture
• Figure 2: A birds eye view semantic segmentation of a traffic scenario at an intersection..