AI Paper Atlas

Research map / Reinforcement learning / Visual and representation RL

VILTA: A VLM-in-the-Loop Adversary for Enhancing Driving Policy Robustness

Qimao Chen, Fang Li, Shaoqing Xu, Zhiyi Lai, Zixun Xie, Yuechen Luo, Shengyin Jiang, Hanbing Li, Long Chen, Bing Wang, Yi Zhang, Zhi-Xin Yang

AAAI 2026 · Reinforcement learning › Visual and representation RL · approach: RL for combinatorial routing · 6 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
train with VLM-in-the-loop trajectory adversary (VILTA) instead of VLM-RLRL driving policy · CARLA Challenging (3 towns total)route completion: 2.10 -> 2.38Improved
train with VILTA instead of CATRL driving policy · CARLA Challenging (3 towns total)crash rate: 1.70 -> 1.03Improved
alternate challenging and normal scenarios (VILTA)RL driving policy · CARLA Normal (3 towns total)route completion: 2.54 -> 2.70Improved
use VILTA instead of VLM-RL on normal scenario speedRL driving policy · CARLA Normalaverage speed: 68.65 -> 65.13Hurt
add trajectory post-processing (B-spline, sigmoid fusion, LQR) (ablation)VILTA · CARLA Challengingroute completion: 0.73 -> 0.77Improved
add following reward (ablation)VILTA · CARLA Challengingroute completion: 0.74 -> 0.77Improved
use Vision-Language-Editing instead of direct VLM generation (ablation)VILTA · CARLA Challengingcrash rate: 0.53 -> 0.50Improved
raise challenging scenario frequency to 1 per 8 normal (from 1 per 2) (ablation)VILTA · CARLA Challengingroute completion: 0.77 -> 0.87Improved
reduce challenging scenarios to 1 per 16 normal (ablation)VILTA · CARLA Challengingroute completion: 0.87 -> 0.85Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy