safety security
The Car Compared Two Stories About Where It Was
Structured narratives turned disagreements between GNSS and independent vehicle sensing into a five-class spoofing detector.
Summary
Structured narratives turned disagreements between GNSS and independent vehicle sensing into a five-class spoofing detector.
The framework converts two independently derived driving states into text-like structured narratives and feeds them to a small language model. Across no-attack, overshoot, stopped, turn-by-turn and wrong-turn cases, the authors report 96.99 percent average accuracy and 97.18 percent F1, with lower latency and memory than their fine-tuned LLM comparisons. Tests on geographically unseen Clemson field data support transfer within the study, not universal resilience to adversarial driving conditions.
Why it matters
Structured narratives turned disagreements between GNSS and independent vehicle sensing into a five-class spoofing detector.
Limits and context
- Tests on geographically unseen Clemson field data support transfer within the study, not universal resilience to adversarial driving conditions.
Key claims
Structured narratives turned disagreements between GNSS and independent vehicle sensing into a five-class spoofing detector.
Qualification: Tests on geographically unseen Clemson field data support transfer within the study, not universal resilience to adversarial driving conditions.
Evidence: source-2026-08-19-010
Sources
- arXiv preprint 2608.17092arXiv · primary research
Corrections
No corrections have been recorded for this story.