YOLOv7-SPD3: A Small Target Detection Algorithm for Multi-Rotor UAV Based on Improved YOLOv7

Lecture notes in electrical engineering, Journal Year: 2025, Volume and Issue: unknown, P. 25 - 34

Published: Jan. 1, 2025

Language: Английский

---

Simplified LSL-Net Architecture for Unmanned Aerial Vehicle Detection in Real-Time

Technologies, Journal Year: 2025, Volume and Issue: 13(5), P. 177 - 177

Published: May 1, 2025

Given the growth of unmanned aerial vehicles (UAVs), their detection has become a recent and complex problem. The literature addressed this problem by applying traditional computer vision algorithms and, more recently, deep learning architectures, which, while proven effective than previous ones, are computationally expensive. In paper, following approach we propose simplified LSL-Net-based architecture for UAV detection. This integrates ability to track detect UAVs using convolutional neural networks. biggest challenge lies in creating model that allows us obtain good results without requiring considerable computational resources. To address problem, built on successful LSL-Net architecture. We introduce dilated convolutions achieve lower-cost with capabilities. Experiments demonstrate our performs well limited resources, reaching 98% accuracy detecting UAVs.

Language: Английский

---

Robustness of Deep-Learning-Based RF UAV Detectors

Sensors, Journal Year: 2024, Volume and Issue: 24(22), P. 7339 - 7339

Published: Nov. 17, 2024

The proliferation of low-cost, small radar cross-section UAVs (unmanned aerial vehicles) necessitates innovative solutions for countering them. Since these typically operate with a radio control link, promising defense technique involves passive scanning the frequency (RF) spectrum to detect UAV signals. This approach is enhanced when integrated machine-learning (ML) and deep-learning (DL) methods. Currently, this field actively researched, various studies proposing different ML/DL architectures competing optimal accuracy. However, there notable gap regarding robustness, which refers detector's ability maintain high accuracy across diverse scenarios, rather than excelling in just one specific test scenario failing others. aspect critical, as inaccuracies detection could lead severe consequences. In work, we introduce new dataset specifically designed robustness. Instead existing extracting data from same pool training data, allowed multiple categories based on channel conditions. Utilizing detectors, found that although coefficient classifiers have outperformed CNNs previous works, our findings indicate image exhibit approximately 40% greater robustness under low signal-to-noise ratio (SNR) Specifically, CNN classifier demonstrated sustained RF conditions not included set, whereas exhibited partial or complete failure depending characteristics.

Language: Английский

---

Hybrid Artificial-Intelligence-Based System for Unmanned Aerial Vehicle Detection, Localization, and Tracking Using Software-Defined Radio and Computer Vision Techniques

Telecom, Journal Year: 2024, Volume and Issue: 5(4), P. 1286 - 1308

Published: Dec. 11, 2024

The proliferation of drones in civilian environments has raised growing concerns about their misuse, highlighting the need to develop efficient detection systems protect public and private spaces. This article presents a hybrid approach for UAV that combines two artificial-intelligence-based methods improve system accuracy. first method uses software-defined radio (SDR) analyze spectrum, employing autoencoders detect drone control signals identify presence these devices. second is computer vision module consisting fixed cameras PTZ camera, which YOLOv10 object algorithm UAVs real time from video sequences. Additionally, this integrates localization tracking algorithm, allowing intruding UAV’s position. Experimental results demonstrate high accuracy, significant reduction false positives both methods, remarkable effectiveness with camera. These findings position proposed as promising solution security applications.

Language: Английский

---