DriveVLM

DriveVLM is a benchmark dataset used to evaluate the vulnerability of visual language models (VLMs) in autonomous driving scenarios, particularly in the context of backdoor attacks.

Papers using DriveVLM (6)

Multimodal Backdoor Attack on VLMs for Autonomous Driving via Graffiti and Cross-Lingual Triggers2026MMDrive: Interactive Scene Understanding Beyond Vision with Multi-representational Fusion2025Natural Reflection Backdoor Attack On Vision Language Model For Autonomous Driving2025TS-VLM: Text-Guided SoftSort Pooling for Vision-Language Models in Multi-View Driving Reasoning2025ReasonDrive: Efficient Visual Question Answering for Autonomous Vehicles with Reasoning-Enhanced Small Vision-Language Models2025Multi-Frame, Lightweight & Efficient Vision-Language Models for Question Answering in Autonomous Driving2024 · 6 cites