Research papers · 2026.03
DriveSOTIF: Advancing Perception SOTIF Through Multimodal Large Language Models
DriveSOTIF: Advancing SOTIF Through Multimodal Large Language Models
Overview
DriveSOTIF fine-tunes multimodal language models on perception-related SOTIF scenarios. Question-answering benchmarks and road cases examine recognition of challenging safety risks, while inference timing is assessed for potential practical use.
Author-supplied source information
Author-supplied information on arXiv.
This work has been accepted to IEEE Transactions on Vehicular Technology. Please refer to the copyright notice for additional information
Version: arXiv v3 — manuscript revision