Saliency-Guided Domain Adaptation for Left-Hand Driving in Autonomous Steering

TL;DR

This paper introduces a saliency-guided domain adaptation method for autonomous steering models to effectively operate in left-hand driving conditions, demonstrating improved accuracy through a novel training pipeline involving flipped data and fine-tuning.

Contribution

It proposes a new training approach combining flipped data pretraining and fine-tuning to enhance domain adaptation for left-hand driving in autonomous vehicles.

Findings

Pretraining on flipped data alone worsens stability.

Fine-tuning after flipped data improves steering accuracy.

Method generalizes across different neural network architectures.

Abstract

Domain adaptation is required for automated driving models to generalize well across diverse road conditions. This paper explores a training method for domain adaptation to adapt PilotNet, an end-to-end deep learning-based model, for left-hand driving conditions using real-world Australian highway data. Four training methods were evaluated: (1) a baseline model trained on U.S. right-hand driving data, (2) a model trained on flipped U.S. data, (3) a model pretrained on U.S. data and then fine-tuned on Australian highways, and (4) a model pretrained on flipped U.S. data and then finetuned on Australian highways. This setup examines whether incorporating flipped data enhances the model adaptation by providing an initial left-hand driving alignment. The paper compares model performance regarding steering prediction accuracy and attention, using saliency-based analysis to measure attention shifts across significant road regions. Results show that pretraining on flipped data alone worsens prediction stability due to misaligned feature representations, but significantly improves adaptation when followed by fine-tuning, leading to lower prediction error and stronger focus on left-side cues. To validate this approach across different architectures, the same experiments were done on ResNet, which confirmed similar adaptation trends. These findings emphasize the importance of preprocessing techniques, such as flipped-data pretraining, followed by fine-tuning to improve model adaptation with minimal retraining requirements.