Human Visual Attention Prediction Boosts Learning & Performance of Autonomous Driving Agents

TL;DR

This paper demonstrates that incorporating human visual attention prediction into autonomous driving models significantly improves their performance by focusing on high-interest regions, reducing control errors by over 25%.

Contribution

The study introduces a novel approach of using human gaze data to generate attention masks, enhancing end-to-end driving models with a dual-branch architecture that outperforms standard methods.

Findings

Dual-branch model reduces control error by 25.5%.

Attention-guided masking improves driving accuracy.

Human gaze prediction enhances autonomous driving performance.

Abstract

Autonomous driving is a multi-task problem requiring a deep understanding of the visual environment. End-to-end autonomous systems have attracted increasing interest as a method of learning to drive without exhaustively programming behaviours for different driving scenarios. When humans drive, they rely on a finely tuned sensory system which enables them to quickly acquire the information they need while filtering unnecessary details. This ability to identify task-specific high-interest regions within an image could be beneficial to autonomous driving agents and machine learning systems in general. To create a system capable of imitating human gaze patterns and visual attention, we collect eye movement data from human drivers in a virtual reality environment. We use this data to train deep neural networks predicting where humans are most likely to look when driving. We then use the outputs of this trained network to selectively mask driving images using a variety of masking techniques. Finally, autonomous driving agents are trained using these masked images as input. Upon comparison, we found that a dual-branch architecture which processes both raw and attention-masked images substantially outperforms all other models, reducing error in control signal predictions by 25.5\% compared to a standard end-to-end model trained only on raw images.