← Back to the feed
Robotics & manufacturingResearch findings

RoboDriveVLM · A Robustness Benchmark for Vision-Language Models in Autonomous Driving

Original published 2025.12.01 14:44 KSTAdded to Blackdeck 2026.11.06 17:21 KST

What has changed?

RoboDriveBench, the first benchmark specifically designed to evaluate robustness of VLM-based autonomous driving systems, is introduced.

Key points

  1. RoboDriveBench, the first robustness benchmark for VLM-based autonomous driving, is introduced.
  2. 11 simulated scenarios including sensor and prompt corruptions are used to evaluate system responses under real-world challenges.
  3. RoboDriveVLM framework enhances robustness by integrating multimodal sensor data into a unified latent space.

Connect it to your work

11 simulated scenarios including sensor and prompt corruptions are used to systematically assess system responses under real-world challenges.

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.

  1. arxivRoboDriveVLM: A Novel Benchmark and Baseline towards Robust Vision-Language Models for Autonomous Driving