Search papers, labs, and topics across Lattice.
This paper introduces the C2E paradigm, which leverages a Multi-to-Single (M2S) agent contrastive knowledge distillation framework to enhance Ego-only 3D object detection in autonomous driving. By integrating a Multi-Level Feature Enhancement module and Auxiliary Point Cloud Reconstruction, the method effectively addresses the limitations of traditional Eo-Perception, such as occlusions and perspective constraints, while maintaining the advantages of Collaborative Perception. Experimental results demonstrate that the M2S framework achieves up to an 8.64% improvement in 3D mean Average Precision (mAP) without incurring additional communication costs.
Ego-only 3D object detection can now achieve significant performance gains without the communication overhead typically associated with multi-agent systems.
LiDAR-based 3D object detection is essential for autonomous driving systems. However, traditional Ego-only Perception (Eo-Perception) suffers from limited perspective and occlusions in a complex outdoor environment, leading to performance bottlenecks. Recently, research on multi-agent Collaborative Perception (Co-Perception) has demonstrated excellent performance, but high communication costs and accumulated pose error hinder its application. To address this, we explore a novel C2E (Co-Perception to Eo-Perception) paradigm through the Multi-to-Single (M2S) agent contrastive knowledge distillation framework. Our M2S framework first designs Multi-Level Feature Enhancement module to provide more stable features, and introduces Auxiliary Point Cloud Reconstruction and Multi-Teacher Contrastive Distillation mechanisms to mitigate domain gaps in point cloud and feature distributions within the C2E paradigm. Benefiting from this, our M2S can retain the excellent performance of collaborative perception while effectively avoiding the drawbacks, such as communication delays and positioning errors. Extensive experiments on the V2XSet, V2V4Real and DAIR-V2X datasets show the effectiveness and generalizability of our M2S framework when combined with the state-of-the-art CoSDH model and other excellent 3D detectors. Our M2S framework can deliver up to a 8.64% improvement in 3D mAP performance without introducing any communication costs.
