SDD-YOLO: A Small-Target Detection Framework for Ground-to-Air Anti-UAV Surveillance with Edge-Efficient Deployment

TL;DR

SDD-YOLO is a specialized small-target detection framework designed for ground-to-air anti-UAV surveillance, featuring high-resolution detection heads and efficient architecture for real-time edge deployment, validated on a large-scale drone dataset.

Contribution

Introduces SDD-YOLO with a high-resolution detection head and advanced training strategies, tailored for micro-UAV detection in G2A scenarios, and constructs a new large-scale drone dataset.

Findings

Achieves 86.0% mAP@0.5 on DroneSOD-30K dataset.

Runs at 226 FPS on NVIDIA RTX 5090, 35 FPS on CPU.

Surpasses baseline YOLOv5n by 7.8 percentage points in accuracy.

Abstract

Detecting small unmanned aerial vehicles (UAVs) from a ground-to-air (G2A) perspective presents significant challenges, including extremely low pixel occupancy, cluttered aerial backgrounds, and strict real-time constraints. Existing YOLO-based detectors are primarily optimized for general object detection and often lack adequate feature resolution for sub-pixel targets, while introducing complexities during deployment. In this paper, we propose SDD-YOLO, a small-target detection framework tailored for G2A anti-UAV surveillance. To capture fine-grained spatial details critical for micro-targets, SDD-YOLO introduces a P2 high-resolution detection head operating at 4 times downsampling. Furthermore, we integrate the recent architectural advancements from YOLO26, including a DFL-free, NMS-free architecture for streamlined inference, and the MuSGD hybrid training strategy with ProgLoss and STAL, which substantially mitigates gradient oscillation on sparse small-target signals. To support our evaluation, we construct DroneSOD-30K, a large-scale G2A dataset comprising approximately 30,000 annotated images covering diverse meteorological conditions. Experiments demonstrate that SDD-YOLO-n achieves a mAP@0.5 of 86.0% on DroneSOD-30K, surpassing the YOLOv5n baseline by 7.8 percentage points. Extensive inference analysis shows our model attains 226 FPS on an NVIDIA RTX 5090 and 35 FPS on an Intel Xeon CPU, demonstrating exceptional efficiency for future edge deployment.