HyDRA: Hybrid Domain-Aware Robust Architecture for Heterogeneous Collaborative Perception

2026-03-25

AI Summary

This work addresses the performance degradation in collaborative perception caused by heterogeneity in model architectures or data distributions among agents. To this end, the authors propose HyDRA, a domain-aware unified fusion framework that dynamically integrates intermediate and late fusion without requiring additional training. HyDRA employs a lightweight domain classifier to identify heterogeneous agents and routes them to a late fusion branch, while leveraging reliable detections from intermediate fusion as spatial anchors to refine the pose graph and mitigate localization errors. The framework enables zero-cost scalability to additional agents, achieving performance on par with state-of-the-art heterogeneous methods while maintaining stable results as the number of agents increases.

Technology Category

• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Multiagent Systems: Adversarial Agents
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

In collaborative perception, an agent's performance can be degraded by heterogeneity arising from differences in model architecture or training data distributions. To address this challenge, we propose HyDRA (Hybrid Domain-Aware Robust Architecture), a unified pipeline that integrates intermediate and late fusion within a domain-aware framework. We introduce a lightweight domain classifier that dynamically identifies heterogeneous agents and assigns them to the late-fusion branch. Furthermore, we propose anchor-guided pose graph optimization to mitigate localization errors inherent in late fusion, leveraging reliable detections from intermediate fusion as fixed spatial anchors. Extensive experiments demonstrate that, despite requiring no additional training, HyDRA achieves performance comparable to state-of-the-art heterogeneity-aware CP methods. Importantly, this performance is maintained as the number of collaborating agents increases, enabling zero-cost scaling without retraining.

Problem

• collaborative perception
• heterogeneity
• model architecture
• data distribution
• agent collaboration

Innovation

• HyDRA
• collaborative perception
• heterogeneous agents
• domain-aware fusion
• pose graph optimization