ScopeMamba YOLO improves small object detection in aerial images
ScopeMamba-YOLO: Widening the Perceptual Scope Inward and Outward for Small Object Detection in Remote Sensing Imagery
Computer Vision and Pattern Recognition
Summary
Detecting tiny objects in drone and satellite images is hard because it requires both clear details and understanding the bigger picture. The authors created ScopeMamba-YOLO, a new technique that looks both closely and broadly without mixing up the two. Their method uses special modules to capture context away from the main image processing stream, improving accuracy with little extra cost. Tests show their approach finds small objects more accurately than some popular existing methods.
What this means in practice
- •For drone software developers: Build better aerial surveillance and mapping software that detects small objects more accurately using remote sensing imagery.
- •For satellite image analysts: Improve automated identification of tiny features in satellite images for environmental monitoring and urban planning.
Authors
Junjie Fan, Yijun Mai, Linduo Wei, Jiayu Rao, Junmin Bao, Qiushi Jin, Guijia Li, Yong Qi
Abstract
Small object detection in unmanned aerial vehicle (UAV) and remote sensing imagery requires preserving high-resolution detail while modeling long-range context. Adding a stride-4 detection level and removing the stride-32 stage benefits tiny targets but weakens peripheral spatial support, whereas directly inserting selective scanning into the main feature path can interfere with weak local cues. We propose ScopeMamba-YOLO, built around an off-path, zero-gated selective-scanning principle that decouples contextual modeling from the convolutional stream. The principle is instantiated by a Cascaded Global-Context Module (CGCM) in the backbone and a Selective-Scan PAN (SS-PAN) in the neck. An Adaptive Multi-scale Strip (AMS) Block reduces the cost of high-resolution feature extraction, while a Scale-Adaptive DFL (SA-DFL) head reallocates distributional support and regression capacity across scales with only 0.008M additional parameters. Controlled experiments show that matched main-path selective scanning reduces mAP50 by 0.98 pp, whereas off-path CGCM improves the final configuration by 0.67 pp over the three-seed no-CGCM mean; operator controls indicate that this gain is not explained by auxiliary branch capacity alone. ERF analysis further shows that the complete context pathway increases the peripheral energy ratio from 0.008 to 0.090 at stride 8. On VisDrone-2019, ScopeMamba-S achieves 50.8% mAP50 with 3.57M parameters, exceeding YOLOv8s by 10.8 pp while using 32% of its parameters; ScopeMamba-M reaches 52.6% mAP50 with 6.48M parameters. Consistent improvements are also observed on AI-TOD, especially for very-tiny and tiny objects.