Diff-KD: Diffusion-based Knowledge Distillation for Collaborative Perception under Corruptions
This addresses robustness issues in autonomous systems' collaborative perception, offering an incremental improvement over existing methods by actively recovering clean semantics from corruptions.
The paper tackles the problem of sensor and communication corruptions undermining multi-agent collaborative perception by introducing Diff-KD, a framework that integrates diffusion-based generative refinement into knowledge distillation, achieving state-of-the-art performance in detection accuracy and calibration robustness on OPV2V and DAIR-V2X datasets under seven corruption types.
Multi-agent collaborative perception enables autonomous systems to overcome individual sensing limits through collective intelligence. However, real-world sensor and communication corruptions severely undermine this advantage. Crucially, existing approaches treat corruptions as static perturbations or passively conform to corrupted inputs, failing to actively recover the underlying clean semantics. To address this limitation, we introduce Diff-KD, a framework that integrates diffusion-based generative refinement into teacher-student knowledge distillation for robust collaborative perception. Diff-KD features two core components: (i) Progressive Knowledge Distillation (PKD), which treats local feature restoration as a conditional diffusion process to recover global semantics from corrupted observations; and (ii) Adaptive Gated Fusion (AGF), which dynamically weights neighbors based on ego reliability during fusion. Evaluated on OPV2V and DAIR-V2X under seven corruption types, Diff-KD achieves state-of-the-art performance in both detection accuracy and calibration robustness.