Domain adaptation affects adversarial robustness in autonomous steering models

Abstract

Domain adaptation is commonly used to transfer autonomous steering models across regions with different road configurations, such as from right-hand to left-hand traffic. While prior work has largely evaluated adaptation using clean target-domain steering accuracy, its effect on adversarial robustness remains insufficiently understood. This study evaluates two steering architectures, PilotNet and ResNet-18, under four adaptation strategies: U.S.-trained, flipped U.S.-trained, U.S.-trained + fine-tuned, and flipped U.S.-trained + fine-tuned. Robustness is assessed under FGSM and PGD attacks across multiple perturbation levels using normalized robustness scores and absolute steering-error metrics. Results show that domain adaptation improves clean target-domain accuracy but affects adversarial robustness in a metric-dependent manner. Fine-tuned models often show degraded normalized adversarial robustness, especially after flipped pretraining, indicating larger relative sensitivity to adversarial perturbations compared with their clean baselines. However, absolute adversarial errors show that fine-tuned models are not always worse than pretrained baselines under attack. These results show that, after domain adaptation, clean accuracy and normalized robustness can diverge, and RS may rank robustness differently from absolute steering-error metrics when clean baselines differ. The contribution of this study is to show that metric choice materially affects robustness conclusions for domain-adapted steering models; therefore, RS remains useful for measuring relative degradation, but should be interpreted alongside absolute steering-error metrics and statistical uncertainty.

HQMSGAN: Enhancing Adversarial Robustness Using Hybrid Quantum-Classical Multi-Scale Generative Adversarial Network
Entropy (Basel, Switzerland)1 Oct 2026Peer-reviewedCC BYFull textEurope PMC
Adversarial Robustness in Fake Quantum Simulators
Marc Maußner, Volker Reers
arXiv1 Oct 2026PreprintarXiv licenceFull textQuantum physics & quantum informationarXiv
On the tradeoff between robustness and geometric accuracy in adversarial training
Rachel Morris, Ryan Murray, Kerrek Stinson
arXiv16 Oct 2026PreprintCC BY 4.0Full textAnalysis & PDEsarXiv
Neural prediction decorrelation reveals that adversarial robustness substantially improves DNN prediction accuracy across the entire human auditory cortex
D. Skrill, J. Feather, S. V. Norman-Haignere
bioRxiv5 Aug 2026PreprintCC BYFull text2 citedNeurosciencebioRxiv
Assessing Adversarial Robustness of Latent Reasoning Models
Shaolong Chen, Ang Li, Mingjie Li +1
arXiv3 Sep 2026PreprintCC BY 4.0Full textStatisticsarXiv

Cite

BibTeX RIS

BibTeX

@article{mehraban2026domain,
  title = {{Domain adaptation affects adversarial robustness in autonomous steering models}},
  author = {Zahra Mehraban and Sebastien Glaser and Michael Milford and Ronald Schroeter},
  journal = {Scientific Reports},
  year = {2026},
  month = oct,
  doi = {10.1038/s41598-026-75207-7},
  url = {https://www.nature.com/articles/s41598-026-75207-7},
  note = {CC BY 4.0},
}

RIS

TY  - JOUR
TI  - Domain adaptation affects adversarial robustness in autonomous steering models
AU  - Zahra Mehraban
AU  - Sebastien Glaser
AU  - Michael Milford
AU  - Ronald Schroeter
PY  - 2026
DA  - 2026/10/12
JO  - Scientific Reports
DO  - 10.1038/s41598-026-75207-7
UR  - https://www.nature.com/articles/s41598-026-75207-7
AB  - Domain adaptation is commonly used to transfer autonomous steering models across regions with different road configurations, such as from right-hand to left-hand traffic. While prior work has largely evaluated adaptation using clean target-domain steering accuracy, its effect on adversarial robustness remains insufficiently understood. This study evaluates two steering architectures, PilotNet and ResNet-18, under four adaptation strategies: U.S.-trained, flipped U.S.-trained, U.S.-trained + fine-tuned, and flipped U.S.-trained + fine-tuned. Robustness is assessed under FGSM and PGD attacks across multiple perturbation levels using normalized robustness scores and absolute steering-error metrics. Results show that domain adaptation improves clean target-domain accuracy but affects adversarial robustness in a metric-dependent manner. Fine-tuned models often show degraded normalized adversarial robustness, especially after flipped pretraining, indicating larger relative sensitivity to adversarial perturbations compared with their clean baselines. However, absolute adversarial errors show that fine-tuned models are not always worse than pretrained baselines under attack. These results show that, after domain adaptation, clean accuracy and normalized robustness can diverge, and RS may rank robustness differently from absolute steering-error metrics when clean baselines differ. The contribution of this study is to show that metric choice materially affects robustness conclusions for domain-adapted steering models; therefore, RS remains useful for measuring relative degradation, but should be interpreted alongside absolute steering-error metrics and statistical uncertainty.
SN  - 2045-2322
ER  -

Abstract

Related

Cite