Vision-Language Models Enhance GNSS Spoofing Detection for Autonomous Vehicles

Researchers have developed a novel method for detecting Global Navigation Satellite Systems (GNSS) spoofing attacks in autonomous vehicles by integrating vision-language models (VLMs) with in-vehicle sensor data. This approach fuses front-camera visual information with readings from sensors like speed and acceleration to identify discrepancies in vehicle maneuvers. The system underwent a three-stage fine-tuning process and was validated on a real-world dataset collected in Tuscaloosa, Alabama, demonstrating significant improvements in detection accuracy, particularly for wrong-turn and stop attacks. AI

IMPACT This research introduces a new defense layer for autonomous vehicles, potentially improving safety and reliability by leveraging VLMs to counter sophisticated navigation attacks.

RANK_REASON Academic paper detailing a new method for GNSS spoofing detection using VLMs. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

• autonomous vehicles
• Global Navigation Satellite Systems
• GNSS spoofing
• Hugging Face
• Tuscaloosa
• vision-language model

• paper
• product
• infra

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →