ArticleScientific reports2025
Effectiveness of YOLO variants for small object detection in SAR images using a new dataset.
Article in Scientific reports, 2025. The graph could read no effect estimate from its abstract, so it casts no vote on the map. Not yet cited in PubMed.
What it found
Each row is one number read from the abstract, on the scale the paper reported it, with its interval. Left of the dashed line favours the treatment, right favours the comparator. Under each row is the sentence it came from. New to these charts? A ten-minute tutorial.
The abstract states no effect estimate the extractor could read, or names no intervention and outcome on the map, so this paper lights no cell and moves no belief. It is still indexed, cited and linked below.
The trial behind it
Trials whose registry record cites this paper, or whose number appears in the abstract. A trial that started after this paper was published is citing it as background, not reporting it.
Neither the registry nor the abstract names a trial number. If this is a trial report, that itself is worth knowing.
Who cites it
0 citing papers in PubMed.
No citing paper in PubMed yet.
Corrections and comments
PubMed lists nothing against this paper. Absence here is not a guarantee, only a check that was made.
Authors and funding
3 authors.
Funding
Abstract
Small object detection in SAR imagery remains challenging due to limited availability of specialized datasets. The article presents a new SAR dataset designed for small-object detection. Due to the absence of publicly available datasets dedicated to vehicle detection on satellite radar imagery, a custom dataset containing 23,644 manually labelled vehicles was created using Capella and ICEYE imagery Also the results of an extensive comparative analysis of three YOLO architectures (versions 7, 8, and 12) in the task of detecting small vehicles in radar imagery were presented. The study also considers the influence of image filtering on detection effectiveness. Experimental results provided new insights into fine-tuning YOLO architectures specifically for detecting small objects in synthetic aperture radar (SAR) images. In addition, the SIVED (SAR Image dataset for VEhicle Detection) dataset (high-resolution airborne imagery) was used in the study. Model performance was tested under various configurations and with Lee, Frost, and GammaMAP filters. Furthermore, a detailed analysis of model stability was performed. The experimental results revealed notable differences in performance among the tested models. The YOLOv8 model achieved the highest detection performance on the SIVED dataset, with an F1-score of 0.958 and mAP@[0.5:0.95] of 0.838 in the unfiltered scenario, along with high stability with respect to changes in threshold parameters. The YOLOv12 model demonstrated its best performance after Lee filtering (F1 score = 0.951, mAP@[0.5:0.95] = 0.774), indicating a greater sensitivity to the quality of the input data. On the contrary, the YOLOv7 model exhibited high sensitivity to changes in confidence thresholds, necessitating precise parameter tuning. The conducted research has shown that YOLOv8 achieves superior detection performance on satellite radar imagery samples despite not incorporating advanced self-attention mechanisms. This work contributes significantly to automatic object detection in radar images, providing practical guidelines for selecting and configuring YOLO models according to the characteristics of the SAR data.
Indexed as
Identifiers
What OpenQuestion holds
Registered trials
Read under generation 80e0d062 · epoch 390. Bibliography from PubMed, PubMed Central and OpenAlex; grants from NIH RePORTER; trial links from ClinicalTrials.gov; estimates, votes and beliefs from the OpenQuestion graph.