Fine-Grained Object Detection with Remote-Sensing Data Using Optimized YOLO-based Models
Bibliographic record
Abstract
Remote sensing data is used in various fields, such as environmental monitoring, urban planning, agriculture, and disaster management. Advances in satellite technology permit the creation of high-quality data, including high-resolution remote-sensing images. However, classical techniques struggle to process and extract information from large quantities and high-resolution remote-sensing data. In this context, with the advancement in computing power, deep learning has become a powerful tool for extracting features and processing remote-sensing data. Despite these advancements, remote-sensing object detection faces challenges related to the diversity of object types, variations in scale and resolution, and the presence of occlusion, which can hinder the accuracy and robustness of detection models. This paper explores YOLO-based architectures for horizontal and oriented bounding box detection in fine-grained object detection using the FAIR1M dataset. FAIR1M provides high-resolution satellite images with 5 object categories and 37 object sub-categories. The best model in horizontal object detection is the YOLOv9e, achieving a mAP50 of 44.6 %. The best model in oriented object detection is a pre-trained model using a custom weighted data loader, achieving a mAP50 of 40.5 %. We further analyze the strengths and limitations of these techniques for fine-grained remote-sensing object detection and highlight the contributions to improving the models' performances depending on the application. We then conclude and give directions for future work.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.002 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".