轻量微调框架提升自动驾驶多模态3D检测鲁棒性
PEFT-DML: Parameter-Efficient Fine-Tuning Deep Metric Learning for Robust Multi-Modal 3D Object Detection in Autonomous Driving
- 用低秩适配与适配器层实现参数高效微调
- 在nuScenes上实现比基线更高精度的检测性能
- 适合传感器故障或组合变化场景的自动驾驶系统
本研究提出PEFT-DML，一种用于自动驾驶中鲁棒多模态3D目标检测的参数高效深度度量学习框架。与传统假设传感器始终可用的模型不同，PEFT-DML将激光雷达、雷达、摄像头、惯导、定位等多元传感器映射到共享隐空间，可在传感器丢失或未见模态组合下仍保持可靠检测能力。通过集成低秩适配（LoRA）与适配器层，该方法在显著提升训练效率的同时，增强了对快速运动、天气变化及域偏移的鲁棒性。在nuScenes基准测试中，其表现优于现有基线模型。
原文摘要 · Abstract (English)
This study introduces PEFT-DML, a parameter-efficient deep metric learning framework for robust multi-modal 3D object detection in autonomous driving. Unlike conventional models that assume fixed sensor availability, PEFT-DML maps diverse modalities (LiDAR, radar, camera, IMU, GNSS) into a shared latent space, enabling reliable detection even under sensor dropout or unseen modality class combinations. By integrating Low-Rank Adaptation (LoRA) and adapter layers, PEFT-DML achieves significant training efficiency while enhancing robustness to fast motion, weather variability, and domain shifts. Experiments on benchmarks nuScenes demonstrate superior accuracy.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。