Lightweight vision language model improves emergency traffic signal control
VLALight: Lightweight Vision-Language-Action Models for Emergency-Aware Traffic Signal Control
Computer Vision and Pattern RecognitionArtificial Intelligence
Summary
Managing traffic signals is important for reducing congestion in cities and helping emergency vehicles get through quickly. The authors created VLALight, a small and fast model that uses multiple camera views and instructions to directly decide traffic light actions. This model skips extra steps that usually slow things down or lose details. Tests showed VLALight significantly cut waiting times for emergency vehicles while working efficiently on regular computers.
What this means in practice
- •For traffic management operators: Improve emergency vehicle response by integrating VLALight for real-time traffic signal decisions that reduce waiting times at intersections.
- •For smart city developers: Use VLALight’s compact vision-language model to enable efficient traffic light systems that adapt to complex intersection views in urban environments.
Authors
Kemou Jiang, Maonan Wang, Xingchen Zou, Jiayue Zhu, Yuhang Fu, Sicheng Wang, Xi Chen, Yirong Chen, Zhiyong Cui
Abstract
Traffic signal control (TSC) is essential for mitigating urban congestion. Recent advances in vision-language models (VLMs) enable richer interpretation of intersection scenes, opening new opportunities for visual-context-aware TSC. However, the loose coupling and repeated information conversion between modules can lead to the loss of fine-grained visual details, while sequential inference introduces substantial latency. To address these limitations, we propose VLALight, a lightweight end-to-end vision-language-action framework that directly maps intersection observations and signal-phase information to discrete signal actions. To handle the multi-view nature of TSC, VLALight combines multiple directional camera views into a unified visual input and uses textual instructions to establish their correspondence with traffic movements and signal phases. This design enables direct action prediction with a compact 0.5 B-parameter model, without intermediate image-to-text descriptions or handcrafted traffic-state representations. Experiments show that VLALight delivers the best emergency-vehicle service of all compared methods, reducing pooled emergency waiting time by 21.1% over the cascaded VLMLight while running in real time on local hardware and generalizing to unseen intersection topologies and traffic-flow patterns.