Enhancing Accuracy and Robustness of Steering Angle Prediction with Attention Mechanism

TL;DR

This paper introduces an attention mechanism into deep neural architectures for autonomous driving, significantly improving steering angle prediction accuracy and robustness against adversarial attacks.

Contribution

It is the first to incorporate attention mechanisms into ResNet and InceptionNet architectures for steering prediction, achieving state-of-the-art results and enhanced robustness.

Findings

Over 6% reduction in steering angle prediction error

Up to 56.09% increase in adversarial robustness

Achieved state-of-the-art MSE performance on benchmark datasets

Abstract

In this paper, our focus is on enhancing steering angle prediction for autonomous driving tasks. We initiate our exploration by investigating two veins of widely adopted deep neural architectures, namely ResNets and InceptionNets. Within both families, we systematically evaluate various model sizes to understand their impact on performance. Notably, our key contribution lies in the incorporation of an attention mechanism to augment steering angle prediction accuracy and robustness. By introducing attention, our models gain the ability to selectively focus on crucial regions within the input data, leading to improved predictive outcomes. Our findings showcase that our attention-enhanced models not only achieve state-of-the-art results in terms of steering angle Mean Squared Error (MSE) but also exhibit enhanced adversarial robustness, addressing critical concerns in real-world deployment. For example, in our experiments on the Kaggle SAP and our created publicly available datasets, attention can lead to over 6% error reduction in steering angle prediction and boost model robustness by up to 56.09%.