arXiv:2507.23473v2 Announce Type: replace 
Abstract: The widespread application of Unmanned Aerial Vehicles (UAVs) has raised serious public safety and privacy concerns, making UAV perception crucial for anti-UAV tasks. However, existing UAV tracking datasets predominantly feature conspicuous objects and lack diversity in scene complexity and attribute representation, limiting their applicability to real-world scenarios. To overcome these limitations, we present the CST Anti-UAV, a new thermal infrared dataset specifically designed for Single Object Tracking (SOT) in Complex Scenes with Tiny UAVs (CST). It contains 220 video sequences with over 240k high-quality bounding box annotations, highlighting two key properties: a significant number of tiny-sized UAV targets and the diverse and complex scenes. To the best of our knowledge, CST Anti-UAV is the first dataset to incorporate complete manual frame-level attribute annotations, enabling precise evaluations under varied challenges. To conduct an in-depth performance analysis for CST Anti-UAV, we evaluate 20 existing SOT methods on the proposed dataset. Experimental results demonstrate that tracking tiny UAVs in complex environments remains a challenge, as the state-of-the-art method achieves only 35.92% state accuracy, much lower than the 67.69% observed on the Anti-UAV410 dataset. These findings underscore the limitations of existing benchmarks and the need for further advancements in UAV tracking research. The CST Anti-UAV benchmark is about to be publicly released, which not only fosters the development of more robust SOT methods but also drives innovation in anti-UAV systems.

تم تقديم مجموعة بيانات CST Anti-UAV لمعالجة المخاوف المتعلقة بالسلامة العامة والخصوصية المرتبطة بالطائرات بدون طيار (UAV). تم تصميم هذه المجموعة الجديدة من البيانات الحرارية تحت الحمراء لتتبع كائن واحد (SOT) في مشاهد معقدة، وتحتوي على 220 تسلسل فيديو وأكثر من 240,000 تعليق عالي الجودة لصناديق التحديد. تبرز الأهداف الصغيرة للطائرات بدون طيار والمشاهد المتنوعة، مما يجعلها أول مجموعة بيانات تتضمن تعليقات يدوية كاملة على مستوى الإطار لتقييمات دقيقة.

Se ha presentado el conjunto de datos CST Anti-UAV para abordar las preocupaciones de seguridad pública y privacidad relacionadas con los Vehículos Aéreos No Tripulados (UAV). Este nuevo conjunto de datos térmico infrarrojo está diseñado para el Seguimiento de Objetos Únicos (SOT) en escenas complejas, con 220 secuencias de video y más de 240,000 anotaciones de cuadros delimitadores de alta calidad. Destaca los objetivos de UAV de tamaño pequeño y escenas diversas, siendo el primer conjunto de datos con anotaciones manuales completas a nivel de fotograma para evaluaciones precisas.

Le jeu de données CST Anti-UAV a été introduit pour répondre aux préoccupations de sécurité publique et de confidentialité liées aux véhicules aériens sans pilote (UAV). Ce nouveau jeu de données infrarouge thermique est conçu pour le suivi d'objets uniques (SOT) dans des scènes complexes, avec 220 séquences vidéo et plus de 240 000 annotations de boîtes englobantes de haute qualité. Il met l'accent sur les cibles UAV de petite taille et des scènes diverses, étant le premier jeu de données avec des annotations d'attributs manuelles complètes au niveau des images pour des évaluations précises.

The CST Anti-UAV dataset has been introduced to address public safety and privacy concerns related to Unmanned Aerial Vehicles (UAVs). This new thermal infrared dataset is designed for Single Object Tracking (SOT) in complex scenes, featuring 220 video sequences and over 240,000 high-quality bounding box annotations. It emphasizes tiny UAV targets and diverse scenes, making it the first dataset with complete manual frame-level attribute annotations for precise evaluations.

CST Anti-UAV: A Thermal Infrared Benchmark for Tiny UAV Tracking in Complex Scenes

Was this article worth reading? Share it

LucidQuery AI

C-THERMAL

Chroma

Attentive AI

Postugc

AIPortalX

Ready to build your own newsroom?