CBDES MoE: Hierarchically Decoupled Mixture-of-Experts for Functional Modules in Autonomous Driving
This work addresses multi-modal perception challenges for autonomous driving systems, presenting a novel modular framework that is incremental in its domain-specific application.
The paper tackles the problem of limited adaptability and capacity in Bird's Eye View perception systems for autonomous driving by proposing a hierarchically decoupled Mixture-of-Experts architecture, achieving a 1.6-point increase in mAP and a 4.1-point improvement in NDS on the nuScenes dataset.
Bird's Eye View (BEV) perception systems based on multi-sensor feature fusion have become a fundamental cornerstone for end-to-end autonomous driving. However, existing multi-modal BEV methods commonly suffer from limited input adaptability, constrained modeling capacity, and suboptimal generalization. To address these challenges, we propose a hierarchically decoupled Mixture-of-Experts architecture at the functional module level, termed Computing Brain DEvelopment System Mixture-of-Experts (CBDES MoE). CBDES MoE integrates multiple structurally heterogeneous expert networks with a lightweight Self-Attention Router (SAR) gating mechanism, enabling dynamic expert path selection and sparse, input-aware efficient inference. To the best of our knowledge, this is the first modular Mixture-of-Experts framework constructed at the functional module granularity within the autonomous driving domain. Extensive evaluations on the real-world nuScenes dataset demonstrate that CBDES MoE consistently outperforms fixed single-expert baselines in 3D object detection. Compared to the strongest single-expert model, CBDES MoE achieves a 1.6-point increase in mAP and a 4.1-point improvement in NDS, demonstrating the effectiveness and practical advantages of the proposed approach.