Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

YOLOBirDrone: Dataset for Bird vs Drone Detection and Classification and a YOLO based enhanced learning architecture

Dapinder Kaur, Neeraj Battish, Arnav Bhavsar, Shashi Poddar

TL;DR

The paper tackles the problem of distinguishing drones from birds in aerial imagery, a challenging task for real-time vision systems due to small object size and appearance similarities. It presents YOLOBirDrone, a YOLOv9-based detector enhanced with AELAN (deformable backbone), MPDA, and RMPDA (multi-scale dual attention) and introduces the BirDrone dataset for robust evaluation. Ablation studies show incremental gains from each module, with the full M6 configuration delivering high precision, strong mAP across IoU thresholds, and a favorable inference time, outperforming state-of-the-art YOLO variants and RT-DETRv2. The work contributes a scalable architecture and a large, challenging dataset to advance safe, real-time drone detection and bird-drone discrimination in practical surveillance scenarios.

Abstract

The use of aerial drones for commercial and defense applications has benefited in many ways and is therefore utilized in several different application domains. However, they are also increasingly used for targeted attacks, posing a significant safety challenge and necessitating the development of drone detection systems. Vision-based drone detection systems currently have an accuracy limitation and struggle to distinguish between drones and birds, particularly when the birds are small in size. This research work proposes a novel YOLOBirDrone architecture that improves the detection and classification accuracy of birds and drones. YOLOBirDrone has different components, including an adaptive and extended layer aggregation (AELAN), a multi-scale progressive dual attention module (MPDA), and a reverse MPDA (RMPDA) to preserve shape information and enrich features with local and global spatial and channel information. A large-scale dataset, BirDrone, is also introduced in this article, which includes small and challenging objects for robust aerial object identification. Experimental results demonstrate an improvement in performance metrics through the proposed YOLOBirDrone architecture compared to other state-of-the-art algorithms, with detection accuracy reaching approximately 85% across various scenarios.

YOLOBirDrone: Dataset for Bird vs Drone Detection and Classification and a YOLO based enhanced learning architecture

TL;DR

The paper tackles the problem of distinguishing drones from birds in aerial imagery, a challenging task for real-time vision systems due to small object size and appearance similarities. It presents YOLOBirDrone, a YOLOv9-based detector enhanced with AELAN (deformable backbone), MPDA, and RMPDA (multi-scale dual attention) and introduces the BirDrone dataset for robust evaluation. Ablation studies show incremental gains from each module, with the full M6 configuration delivering high precision, strong mAP across IoU thresholds, and a favorable inference time, outperforming state-of-the-art YOLO variants and RT-DETRv2. The work contributes a scalable architecture and a large, challenging dataset to advance safe, real-time drone detection and bird-drone discrimination in practical surveillance scenarios.

Abstract

The use of aerial drones for commercial and defense applications has benefited in many ways and is therefore utilized in several different application domains. However, they are also increasingly used for targeted attacks, posing a significant safety challenge and necessitating the development of drone detection systems. Vision-based drone detection systems currently have an accuracy limitation and struggle to distinguish between drones and birds, particularly when the birds are small in size. This research work proposes a novel YOLOBirDrone architecture that improves the detection and classification accuracy of birds and drones. YOLOBirDrone has different components, including an adaptive and extended layer aggregation (AELAN), a multi-scale progressive dual attention module (MPDA), and a reverse MPDA (RMPDA) to preserve shape information and enrich features with local and global spatial and channel information. A large-scale dataset, BirDrone, is also introduced in this article, which includes small and challenging objects for robust aerial object identification. Experimental results demonstrate an improvement in performance metrics through the proposed YOLOBirDrone architecture compared to other state-of-the-art algorithms, with detection accuracy reaching approximately 85% across various scenarios.
Paper Structure (15 sections, 8 equations, 4 figures, 5 tables)

This paper contains 15 sections, 8 equations, 4 figures, 5 tables.

Table of Contents

  1. INTRODUCTION
  2. RELATED WORK
  3. YOLO-based architectures
  4. Attention mechanisms
  5. PROPOSED METHODOLOGY
  6. YOLOv9
  7. YOLOBirDrone
  8. AELAN
  9. MPDA and RMPDA
  10. EXPERIMENTATION AND RESULTS ANALYSIS
  11. Dataset details
  12. Ablation Study
  13. Detection Results
  14. Comparison of the proposed YOLOBirDrone with other State-of-the-art (SOTA) methods
  15. CONCLUSION

Figures (4)

  • Figure 1: Proposed YOLOBirDrone Architecture
  • Figure 2: Proposed AELAN sub-architecture
  • Figure 3: MPDA and RMPDA Modules
  • Figure 4: Detection results of proposed YOLOBirDrone