J. Eur. Opt. Society-Rapid Publ. 2026, 22, 43 Ó The Author( s), published by EDP Sciences, 2026 https:// doi. org / 10.1051 / jeos / 2026037 Available online at: https:// jeos. edpsciences. org
Journal of the European Optical Society-Rapid Publications
RESEARCH ARTICLE
SIM-AIR dataset and YOLO-KMM model for air-to-air infrared small target detection
Luyi Zhang, Limin Liu, Chaowen Zheng, Haojie Yang, and Qiang Fu * Shijiazhuang Campus, Army Engineering University of PLA, Shijiazhuang 050003, PR China
Received 25 January 2026 / Accepted 8 April 2026
Abstract. To address the lack of dedicated datasets for infrared detection of small UAVs in air-to-air scenarios, this paper first constructs the self-built SIM-AIR dataset covering complex scenarios, and then proposes YOLO-KMM an efficient YOLOv11-based object detection model tailored to the dataset’ s small-target characteristics and deployment requirements; collected by an UAV equipped with an infrared thermal imager, the SIM-AIR dataset consists of 3,993 precisely annotated images across four weather conditions: sunny, cloudy, snowy, and hazy, where 99.7 % of the targets are ultra-small objects and their width < 40 pixels, with an average size of 11.2 6.6 pixels, including complex scenarios such as“ dark targets” in snowy weather and low signal-to-noise ratio( SNR) in haze, which fully simulate real-world detection challenges. To tackle the issues of sparse small-target features and strong background interference, YOLO-KMM integrates the C2KD feature enhancement module and C3K2-MU lightweight detection head, forming a dual-optimized architecture of“ feature enhancement – efficient detection”: the C2KD module captures weak small-target features and suppresses noise via cross-scale fusion and attention mechanisms, while the C3K2-MU module adopts grouped convolution and depthwise separable convolution to reduce the number of parameters while preserving feature representation capability. Experiments on the SIM-AIR dataset show that YOLO-KMM achieves an mAP 50 of 88.2 %. This is 7.8 % points higher than the baseline YOLOv11, with a precision of 94.0 % and recall of 74.3 %, reduces the small-target missed detection rate by 12.5 %, and maintains an inference speed of 246.18 FPS, 2.3M parameters, and 5.4 GFLOPs of computation; compared with YOLOv5 / 8 / 12, the model achieves a better balance among accuracy, speed, and complexity, verifying the practicality and challenge of the SIM-AIR dataset and providing an efficient solution for air-to-air small-target infrared detection.
Keywords: Air-to-air infrared detection, Small UAV target, SIM-AIR dataset, YOLO-KMM model, Feature enhancement, Lightweight object detection.
1 Introduction
The detection of small air-to-air drones has become a crucial technological necessity in the fields of civilian security, transportation, and emergency rescue. In civil aviation, the International Civil Aviation Organization( ICAO) recorded over 1,200 near-miss incidents between drones and aircraft in 2023, 68 % of which occurred in low-altitude airspace where conventional radar fails, with single incidents potentially causing economic losses of several million dollars. In emergency rescue scenarios, collisions caused by complex weather conditions such as smoke, rain, or snow often result in monitoring interruptions and rescue delays. With the development of urban air mobility( UAM), the global civilian drone fleet is expected to exceed 20 million units by
* Corresponding author: fu _ qiang @ aeu. edu. cn
2026, creating an urgent demand for real-time collision avoidance technology in low-altitude airspace.
Infrared detection has become a core solution due to its all-weather operational capability, but existing technologies struggle with challenges such as ultra-small targets, low signal-to-noise ratios, and limited UAV platform resources, and also lack dedicated datasets. To address this, this paper constructs the SIM-AIR dataset and proposes the YOLO- KMM model, specifically tackling these practical challenges and providing a practical technical solution for airborne infrared small target detection.
As a core research direction in computer vision, object detection focuses on fast and accurate target localization and category recognition, and has been widely applied in scenarios such as security monitoring, industrial quality inspection, and intelligent transportation. In recent years, single-stage object detectors represented by the YOLO series have become the preferred solution for real-time
This is an Open Access article distributed under the terms of the Creative Commons Attribution License( https:// creativecommons. org / licenses / by / 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.