SL-YOLO提升无人机小目标检测精度，同时更轻更快。
SL-YOLO: A Stronger and Lighter Drone Target Detection Model
- 设计跨尺度特征融合网络HEPAN，增强小目标特征提取能力。
- 在VisDrone2019上[email protected]达46.9%，比原模型提升3.9个百分点。
- 参数减少至9.6M，推理速度达132 FPS，适合边缘设备部署。
在复杂场景中检测小目标（如无人机拍摄图像）极具挑战性，因难以捕捉其复杂特征。尽管YOLO系列在大目标检测中表现优异，但对小目标效果不佳。为此，本文提出SL-YOLO（Stronger and Lighter YOLO），突破小目标检测瓶颈。创新设计层级扩展路径聚合网络（HEPAN），实现卓越的跨尺度特征融合，显著提升检测精度。同时，引入C2fDCB轻量化模块与SCDown下采样模块，在不损失检测性能的前提下，将模型参数从11.1M降至9.6M，计算量大幅降低。在VisDrone2019数据集上的实验表明，[email protected]由43.0%提升至46.9%，[email protected]:0.95由26.0%增至28.9%，推理速度达132 FPS，为资源受限环境下的实时小目标检测提供了高效解决方案。
原文摘要 · Abstract (English)
Detecting small objects in complex scenes, such as those captured by drones, is a daunting challenge due to the difficulty in capturing the complex features of small targets. While the YOLO family has achieved great success in large target detection, its performance is less than satisfactory when faced with small targets. Because of this, this paper proposes a revolutionary model SL-YOLO (Stronger and Lighter YOLO) that aims to break the bottleneck of small target detection. We propose the Hierarchical Extended Path Aggregation Network (HEPAN), a pioneering cross-scale feature fusion method that can ensure unparalleled detection accuracy even in the most challenging environments. At the same time, without sacrificing detection capabilities, we design the C2fDCB lightweight module and add the SCDown downsampling module to greatly reduce the model's parameters and computational complexity. Our experimental results on the VisDrone2019 dataset reveal a significant improvement in performance, with [email protected] jumping from 43.0% to 46.9% and [email protected]:0.95 increasing from 26.0% to 28.9%. At the same time, the model parameters are reduced from 11.1M to 9.6M, and the FPS can reach 132, making it an ideal solution for real-time small object detection in resource-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。