Energy & materialsResearch findings
End-to-End Autonomous Driving Vulnerability via Module-wise Noise Injection
Original published 2024.09.12 11:19 KSTAdded to Blackdeck 2026.11.12 22:32 KST
What has changed?
A universal attack scheme using module-wise noise injection is proposed
Key points
- Comprehensive adversarial security research is conducted on modular end-to-end autonomous driving models
- Vulnerabilities in the inference process across modules are systematically analyzed
- Large-scale experiments on full-stack models demonstrate superior attack performance over prior methods
Connect it to your work
A universal attack scheme using module-wise noise injection is proposed
What remains unproven
This research note is based on an abstract; check the original for test conditions and application scope.
Originals used for this note
Links open the original publisher in a new tab.