Development and testing of an image transformer for explainable autonomous driving systems

TL;DR

This paper introduces an explainable transformer-based autonomous driving system that improves interpretability and performance over benchmarks, enhancing user trust and diagnostic capabilities in self-driving cars.

Contribution

The paper presents a novel transformer-based model for autonomous driving that provides explanations for its decisions, addressing interpretability issues in deep learning models.

Findings

Achieves higher accuracy in driving actions and explanations compared to benchmarks

Exhibits lower computational cost than existing models

Provides effective global attention visualization for interpretability

Abstract

In the last decade, deep learning (DL) approaches have been used successfully in computer vision (CV) applications. However, DL-based CV models are generally considered to be black boxes due to their lack of interpretability. This black box behavior has exacerbated user distrust and therefore has prevented widespread deployment DLCV models in autonomous driving tasks even though some of these models exhibit superiority over human performance. For this reason, it is essential to develop explainable DL models for autonomous driving task. Explainable DL models can not only boost user trust in autonomy but also serve as a diagnostic approach to identify anydefects and weaknesses of the model during the system development phase. In this paper, we propose an explainable end-to-end autonomous driving system based on "Transformer", a state-of-the-art (SOTA) self-attention based model, to map visual features from images collected by onboard cameras to guide potential driving actions with corresponding explanations. The model achieves a soft attention over the global features of the image. The results demonstrate the efficacy of our proposed model as it exhibits superior performance (in terms of correct prediction of actions and explanations) compared to the benchmark model by a significant margin with lower computational cost.