DroBoost: An Intelligent Score and Model Boosting Method for Drone Detection

TL;DR

DroBoost introduces an enhanced drone detection method that combines diverse datasets, synthetic samples, classification, and advanced scoring to improve detection accuracy and resilience, winning first place in a competitive challenge.

Contribution

It presents a novel integrated approach that improves drone detection by combining multiple data sources, a classification component, and an advanced scoring algorithm, surpassing previous methods.

Findings

Achieved first place in the Drone vs. Bird Challenge.

Enhanced detection confidence through a new scoring algorithm.

Improved resilience and accuracy in drone detection under challenging conditions.

Abstract

Drone detection is a challenging object detection task where visibility conditions and quality of the images may be unfavorable, and detections might become difficult due to complex backgrounds, small visible objects, and hard to distinguish objects. Both provide high confidence for drone detections, and eliminating false detections requires efficient algorithms and approaches. Our previous work, which uses YOLOv5, uses both real and synthetic data and a Kalman-based tracker to track the detections and increase their confidence using temporal information. Our current work improves on the previous approach by combining several improvements. We used a more diverse dataset combining multiple sources and combined with synthetic samples chosen from a large synthetic dataset based on the error analysis of the base model. Also, to obtain more resilient confidence scores for objects, we introduced a classification component that discriminates whether the object is a drone or not. Finally, we developed a more advanced scoring algorithm for object tracking that we use to adjust localization confidence. Furthermore, the proposed technique won 1st Place in the Drone vs. Bird Challenge (Workshop on Small-Drone Surveillance, Detection and Counteraction Techniques at ICIAP 2021).