Acoustic-Based UAV Detection Using Late Fusion of Deep Neural Networks

Casabianca, Pietro; Zhang, Yu

doi:10.3390/drones5030054

Cited by 31 publications

(16 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…On the other hand, acoustic-based algorithms has proven to be promising and feasible for short range drones surveillance [34]- [36]. In [12], support vector machines (SVM)based method has been used for UAV sounds classification among birds, airplanes or thunderstorms.…”

Section: ) Other Detection Techniquesmentioning

confidence: 99%

Malicious UAV Detection Over Rician Fading Channel: Performance Analysis

Awad,

Al-Dharrab

2024

IEEE Access

View full text Add to dashboard Cite

Unmanned aerial vehicles (UAVs) are predicted to be widely used in both military and civilian sectors in the coming years due to their high mobility, low cost, and enhancement of the line-of-sight (LoS) conditions in non-terrestrial networks. Nevertheless, this raises some security issues if they are manipulated to cause security threats in restricted locations or to breach user privacy. In order to detect malicious UAVs, radio frequency (RF)-based approaches are adopted to detect ambient RF signals, which can be accomplished with inexpensive RF sensors under both LoS and, in particular, non-line-of-sight (NLoS) conditions. In this paper, we propose a passive detection technique based on received signal strength (RSS), and derive analytical expressions on the detection and false alarm probabilities considering realistic airto-ground (A2G) channel conditions. A novel low-complexity suboptimal detector is also proposed and its performance is compared to the optimal detection. Monte Carlo simulations are used to confirm the accuracy of the derived expressions under the aforementioned channel conditions. Our mathematical framework, analytical derivations, and simulation results reveal that the sensing node can achieve an accuracy of 0.9 under LoS scenarios, where the NLoS conditions cause some challenges in the accuracy of detection. The low-complexity suboptimal detector for urban and suburban environments has close performance compared to the optimal detection.INDEX TERMS Air-to-ground (A2G), log-likelihood ratio, receiver operating characteristic (ROC), sufficient statistic.

show abstract

Section: ) Other Detection Techniquesmentioning

confidence: 99%

Malicious UAV Detection Over Rician Fading Channel: Performance Analysis

Awad,

Al-Dharrab

2024

IEEE Access

View full text Add to dashboard Cite

show abstract

“…The results show that the classifier can precisely distinguish the UAVs in some scenarios [ 33 ]. Another example of using deep learning methods to detect UAVs with acoustic signals is shown in [ 34 ]. In this paper, there is a comparison among Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs) and Convolutional Recurrent Neural Networks (CRNNs) using melspectrogram features.…”

Section: Review Of the State-of-the-art Technologymentioning

confidence: 99%

Review and Simulation of Counter-UAS Sensors for Unmanned Traffic Management

Besada

Campaña

Carramiñana

et al. 2021

Sensors

View full text Add to dashboard Cite

Noncollaborative surveillance of airborne UAS (Unmanned Aerial System) is a key enabler to the safe integration of UAS within a UTM (Unmanned Traffic Management) ecosystem. Thus, a wide variety of new sensors (known as Counter-UAS sensors) are being developed to provide real-time UAS tracking, ranging from radar, RF analysis and image-based detection to even sound-based sensors. This paper aims to discuss the current state-of-the art technology in this wide variety of sensors (both academically and commercially) and to propose a set of simulation models for them. Thus, the review is focused on identifying the key parameters and processes that allow modeling their performance and operation, which reflect the variety of measurement processes. The resulting simulation models are designed to help evaluate how sensors’ performances affect UTM systems, and specifically the implications in their tracking and tactical services (i.e., tactical conflicts with uncontrolled drones). The simulation models cover probabilistic detection (i.e., false alarms and probability of detection) and measurement errors, considering equipment installation (i.e., monostatic vs. multistatic configurations, passive sensing, etc.). The models were integrated in a UTM simulation platform and simulation results are included in the paper for active radars, passive radars, and acoustic sensors.

show abstract

“…Nonetheless, the availability of drones has posed significant privacy and secrecy dilemma. Moreover, to emphasize the significance of the problem, we noticed security threats from the uncontrolled usage of UAVs that severely damaged the infrastructure [4]. Drones were initially developed for defense and counterinsurgency and controlled by the aerospace and defense industries.…”

Section: Introductionmentioning

confidence: 99%

TransLearn-YOLOx: Improved-YOLO with Transfer Learning for Fast and Accurate Multiclass UAV Detection

Khan¹,

Dil²,

Misbah³

et al. 2023

Preprint

View full text Add to dashboard Cite

The emergence of unmanned aerial vehicles (UAVs) raised multiple concerns, given their potential for malicious misuse in unlawful acts Vision-based counter-UAV applications offer a reliable solution compared to acoustic and radio frequency-based solutions because of their high detection accuracy in diverse weather conditions. The existing solutions work well on trained datasets, but their accuracy is relatively low for real-time detection. In this paper, we model deep learning-empowered solutions to improve the multi-class UAV's classification performance using single-shot object detection algorithms YOLOv5 and YOLOv7. The transfer learning is employed for performance improvement and rapid training with improved results. We customized a multi-class dataset containing multi-rotor, fixed-wing, and single-rotor UAVs in challenging weather conditions. Experiments show that the integration of transfer learning has achieved good results, with an overall best average-classification precision of 94\%, an average recall of 93.1\%, a mAP$@$0.5 average of 95.3\%, and an average F1 score of 92.33\%. The dataset and code are available as an open source: https://github.com/ZeeshanKaleem/YOLOV5-Large-vs-YOLOV7.git

show abstract

Acoustic-Based UAV Detection Using Late Fusion of Deep Neural Networks

Cited by 31 publications

References 23 publications

Malicious UAV Detection Over Rician Fading Channel: Performance Analysis

Malicious UAV Detection Over Rician Fading Channel: Performance Analysis

Review and Simulation of Counter-UAS Sensors for Unmanned Traffic Management

TransLearn-YOLOx: Improved-YOLO with Transfer Learning for Fast and Accurate Multiclass UAV Detection

Contact Info

Product

Resources

About