A Deep Learning Approach for Drone Detection and Classification Using Radar and Camera Sensor Fusion
Bibliographic record
Abstract
With the growth of Unmanned Aircraft Systems (UAS) technology and the increasing misuse of small UAS (sUAS), the importance of a reliable method for detecting and classifying aircraft from other flying objects has become apparent. The current approaches for detecting and classifying aircraft and other flying objects are primarily based on solutions that rely on a single sensor, either visual data features or micro-Doppler extraction from radar data. However, these methods may have limitations when it comes to detecting objects at greater distances or in challenging weather conditions. To address the problem, the paper proposes a joint classification network based on radar and camera fusion. The radar network extracts the Spatio temporal features from the radar track and the camera network extracts the deep, complex features from the image. A synchronized radar and camera data is established using multiple field trials during different times of the year. The radar classification using a combination of IMM filters and RNN, the camera detection and classification using YOLOv5, and the combined joint classification network are evaluated on the field dataset. The experimental results greatly increase the classification performance for drones and birds, respectively, to 98% and 94%. This is especially true in situations when a single sensor would struggle to offer reliable classification. The system can accurately classify drones while reducing false alarms caused by other objects, such as birds.
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 | 0.001 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
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".