HyDRA: Hybrid Domain-Aware Robust Architecture for Heterogeneous Collaborative Perception
This addresses robustness and scalability issues in collaborative perception for autonomous systems, though it is incremental as it builds on existing fusion methods.
The paper tackles performance degradation in collaborative perception due to agent heterogeneity by proposing HyDRA, a unified pipeline that integrates intermediate and late fusion with domain-aware classification and anchor-guided optimization, achieving performance comparable to state-of-the-art methods without additional training and maintaining it as agent numbers increase.
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.