针对无人机图像中微小目标检测难题，提出轻量级高效检测框架
DroneScan-YOLO: Redundancy-Aware Lightweight Detection for Tiny Objects in UAV Imagery
- 采用1280x1280高分辨率输入并引入步长为4的轻量检测分支
- 在VisDrone2019上实现55.3% mAP@50，tiny目标检测提升超180%
- 动态滤波剪枝+自适应损失函数，兼顾精度与96.7FPS实时性
无人机影像中的空中目标检测面临微小目标普遍、环境恶劣及计算资源受限等挑战。标准YOLO检测器存在多重缺陷：最小下采样步长8像素导致小于32像素的目标几乎无法检测，CIoU损失对无重叠的微小框梯度为零，且模型存在显著滤波冗余。本文提出DroneScan-YOLO，通过四项协同设计解决上述问题：(1) 输入分辨率提升至1280x1280以增强空间细节；(2) RPA-Block，基于懒惰余弦相似度更新的动态滤波剪枝机制，含10轮预热期；(3) MSFD，stride为4的轻量级P2检测分支，仅增加114,592参数（+1.1%）；(4) SAL-NWD，融合归一化Wasserstein距离与尺寸自适应CIoU加权的混合损失，集成至YOLOv8的任务对齐分配流程。在VisDrone2019-DET数据集上，DroneScan-YOLO达到55.3% mAP@50和35.6% mAP@50-95，分别比YOLOv8s基线提升+16.6和+12.3点，召回率从0.374提升至0.518，推理速度维持96.7 FPS，参数仅增加4.1%。对微小目标类别增益尤为显著：自行车类AP@50从0.114提升至0.328（+187%），遮阳车篷三轮车从0.156升至0.237（+52%）。
原文摘要 · Abstract (English)
Aerial object detection in UAV imagery presents unique challenges due to the high prevalence of tiny objects, adverse environmental conditions, and strict computational constraints. Standard YOLO-based detectors fail to address these jointly: their minimum detection stride of 8 pixels renders sub-32px objects nearly undetectable, their CIoU loss produces zero gradients for non-overlapping tiny boxes, and their architectures contain significant filter redundancy. We propose DroneScan-YOLO, a holistic system contribution that addresses these limitations through four coordinated design choices: (1) increased input resolution of 1280x1280 to maximize spatial detail for tiny objects, (2) RPA-Block, a dynamic filter pruning mechanism based on lazy cosine-similarity updates with a 10-epoch warm-up period, (3) MSFD, a lightweight P2 detection branch at stride 4 adding only 114,592 parameters (+1.1%), and (4) SAL-NWD, a hybrid loss combining Normalized Wasserstein Distance with size-adaptive CIoU weighting, integrated into YOLOv8's TaskAligned assignment pipeline. Evaluated on VisDrone2019-DET, DroneScan-YOLO achieves 55.3% mAP@50 and 35.6% mAP@50-95, outperforming the YOLOv8s baseline by +16.6 and +12.3 points respectively, improving recall from 0.374 to 0.518, and maintaining 96.7 FPS inference speed with only +4.1% parameters. Gains are most pronounced on tiny object classes: bicycle AP@50 improves from 0.114 to 0.328 (+187%), and awning-tricycle from 0.156 to 0.237 (+52%).
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。