V2X-DG: Domain Generalization for Vehicle-to-Everything Cooperative Perception

2025-03-19

AI Summary

This work addresses the domain generalization (V2X-DG) problem for LiDAR-based 3D object detection in V2X cooperative perception. We establish the first cross-dataset, target-domain-free benchmark—covering OPV2V and three other major open-source datasets—enabling rigorous evaluation without access to target-domain data during training. To tackle unknown collaboration patterns across domains, we propose Collaborative Mixup for Augmentation and Generalization (CMAG), which explicitly models heterogeneous agent interactions. Additionally, we introduce Cooperative Feature Consistency (CFC) regularization to enhance robustness of cross-domain feature representations. Our method integrates intermediate- and early-stage feature alignment with domain-generalization regularization, achieving significant performance gains on multiple unseen domains while preserving high accuracy on source domains. This is the first systematic solution for V2X domain generalization, providing foundational technical support for real-world deployment of vehicle-infrastructure-cloud cooperative perception systems.

Technology Category

• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Computer Vision: Large Vision Models
• Application Domains: Mobility, Driving & Flight

Application Category

• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Systems and Infrastructure for Web, Mobile and WoT: Web applications in cross-disciplinary domains and verticals such as mixed reality, smart cities, and digital health
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

LiDAR-based Vehicle-to-Everything (V2X) cooperative perception has demonstrated its impact on the safety and effectiveness of autonomous driving. Since current cooperative perception algorithms are trained and tested on the same dataset, the generalization ability of cooperative perception systems remains underexplored. This paper is the first work to study the Domain Generalization problem of LiDAR-based V2X cooperative perception (V2X-DG) for 3D detection based on four widely-used open source datasets: OPV2V, V2XSet, V2V4Real and DAIR-V2X. Our research seeks to sustain high performance not only within the source domain but also across other unseen domains, achieved solely through training on source domain. To this end, we propose Cooperative Mixup Augmentation based Generalization (CMAG) to improve the model generalization capability by simulating the unseen cooperation, which is designed compactly for the domain gaps in cooperative perception. Furthermore, we propose a constraint for the regularization of the robust generalized feature representation learning: Cooperation Feature Consistency (CFC), which aligns the intermediately fused features of the generalized cooperation by CMAG and the early fused features of the original cooperation in source domain. Extensive experiments demonstrate that our approach achieves significant performance gains when generalizing to other unseen datasets while it also maintains strong performance on the source dataset.

Problem

• Addresses domain generalization in V2X cooperative perception.
• Improves model generalization across unseen domains.
• Proposes CMAG and CFC for robust feature learning.

Innovation

• Cooperative Mixup Augmentation for generalization
• Cooperation Feature Consistency for alignment
• Training on source domain for unseen domains