用自监督模型提升工程领域小样本目标检测性能
DINO-YOLO: Self-Supervised Pre-training for Data-Efficient Object Detection in Civil Engineering Applications
- 融合YOLOv12与DINOv3，在输入和中间层注入自监督特征
- 小数据集上检测精度提升超12%，最高达88.6%改进
- 适合工地安全监控等数据少但需实时检测的场景
土木工程中的目标检测受限于特定领域标注数据稀少。本文提出DINO-YOLO，一种结合YOLOv12与DINOv3自监督视觉变换器的混合架构，实现数据高效检测。DINOv3特征被战略性地融入输入预处理（P0）和主干网络中段增强（P3）。实验验证显示：隧道衬砌裂缝检测（648张图）提升12.4%，施工安全装备识别（1000张图）提升13.7%，KITTI数据集（7000张图）提升88.6%，同时保持实时推理速度（30-47 FPS）。五种YOLO尺度与九种DINOv3变体的系统消融表明，中等规模架构在双路径融合（DualP0P3）下表现最优（[email protected]为55.77%），小规模需三路融合（Triple Integration，53.63%）。推理延迟增加2-4倍（21-33毫秒对比基线8-16毫秒），在NVIDIA RTX 5090上仍可接受。DINO-YOLO在小于1万图像的土木工程数据集上达到当前最佳性能，兼顾计算效率，为数据受限环境下的施工安全监测与基础设施巡检提供实用方案。
原文摘要 · Abstract (English)
Object detection in civil engineering applications is constrained by limited annotated data in specialized domains. We introduce DINO-YOLO, a hybrid architecture combining YOLOv12 with DINOv3 self-supervised vision transformers for data-efficient detection. DINOv3 features are strategically integrated at two locations: input preprocessing (P0) and mid-backbone enhancement (P3). Experimental validation demonstrates substantial improvements: Tunnel Segment Crack detection (648 images) achieves 12.4% improvement, Construction PPE (1K images) gains 13.7%, and KITTI (7K images) shows 88.6% improvement, while maintaining real-time inference (30-47 FPS). Systematic ablation across five YOLO scales and nine DINOv3 variants reveals that Medium-scale architectures achieve optimal performance with DualP0P3 integration (55.77% [email protected]), while Small-scale requires Triple Integration (53.63%). The 2-4x inference overhead (21-33ms versus 8-16ms baseline) remains acceptable for field deployment on NVIDIA RTX 5090. DINO-YOLO establishes state-of-the-art performance for civil engineering datasets (<10K images) while preserving computational efficiency, providing practical solutions for construction safety monitoring and infrastructure inspection in data-constrained environments.
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。