LACO: Adaptive Latent Communication for Collaborative Driving

2026-05-21

AI Summary

This work addresses the challenges of high latency and information loss in linguistic communication, as well as agent identity confusion arising from latent-space fusion in cooperative driving. To overcome these issues, the authors propose LACO, a training-free, adaptive latent-space communication framework that enables efficient, low-latency collaborative decision-making through Iterative Latent-state Deduction (ILD), Cross-temporal Hierarchical Significance Attribution (CHSA), and Structured Semantic Knowledge Distillation (SSKD). LACO is the first method to explicitly identify and resolve identity confusion in latent-space communication. Evaluated in closed-loop CARLA experiments, it significantly reduces both communication and inference latency while maintaining superior cooperative driving performance.

Abstract

Collaborative driving aims to improve safety and efficiency by enabling connected vehicles to coordinate under partial observability. Recent approaches have evolved from sharing visual features for perception to exchanging language-based reasoning through foundation models for behavioral coordination. Though communicating in language provides intuitive information, it introduces two challenges: high latency caused by autoregressive decoding and information loss caused by compressing rich internal representations into discrete tokens. To address these challenges, we analyze latent communication in collaborative driving under inherent limitations of multi-agent settings. Our analysis reveals agent identity confusion, where direct fusion of latent states entangles decision representations across vehicles. Motivated by this, we propose LACO, a training-free \textbf{LA}tent \textbf{CO}mmunication paradigm that seamlessly adapts pretrained driving models to collaborative settings. LACO introduces Iterative Latent Deliberation (ILD) for latent reasoning, Cross-Horizon Saliency Attribution (CHSA) for communication-efficient information selection, and Structured Semantic Knowledge Distillation (SSKD) to stabilize ego-centric decision making. Closed-loop experiments in CARLA show that LACO notably reduces communication and inference latency while maintaining strong collaborative driving performance.

Problem

• collaborative driving
• latent communication
• agent identity confusion
• communication latency
• information loss

Innovation

• Latent Communication
• Collaborative Driving
• Training-Free Adaptation
• Iterative Latent Deliberation
• Cross-Horizon Saliency Attribution