Abstract
With the widespread application of drone in commercial and industrial fields, drone detection has received increasing attention in public safety and others. However, due to various appearance of small-size drones, changeable and complex environments, and limited memory resources of edge computing devices, drone detection remains a challenging task nowadays. Although deep convolutional neural network (CNN) has shown powerful performance in object detection in recent years, most existing CNN-based methods cannot balance detection performance and model size well. To solve the problem, we develop a drone detection network with tiny iterative backbone named TIB-Net. In this network, we propose a structure called cyclic pathway, which enhances the capability to extract effective features of small object, and integrate it into existing efficient method Extremely Tiny Face Detector (EXTD). This method not only significantly improves the accuracy of drone detection, but also keeps the model size at an acceptable level. Furthermore, we integrate spatial attention module into our network backbone to emphasize information of small object, which can better locate small-size drone and further improve detection performance. In addition, we present massive manual annotations of object bounding boxes for our collected 2860 drone images as a drone benchmark dataset, which is now publicly available 1 . In this work, we conduct a series of experiments on our collected dataset to evaluate TIB-Net, and the result shows that our proposed method achieves mean average precision of 89.2% with model size of 697.0KB, which achieves the state-of-the-art results compared with existing methods.
Highlights
With the rapid development of drone technology, more and more drones have been utilized for commercial and other purposes, which threaten public safety and air traffic to some extent [1]
The main contributions of this work can be summarized as follows: (1) We propose a detection network with tiny iterative network named TIB-Net for drone detection, and the network integrates with cyclic pathway structure to obtain more texture and contour information, which can significantly improve the accuracy of drone detection, especially to smallsize drone
We first verify the effectiveness of the TIB-Net’s components, we compared it with other object detection methods on accuracy, speed, and model size
Summary
With the rapid development of drone technology, more and more drones have been utilized for commercial and other purposes, which threaten public safety and air traffic to some extent [1]. Compared with the traditional methods, the CNN-based methods can extract much richer high-level representation features by transmitting feature maps through a series of alternating network layers, which promote the development of object detection greatly. Most of these CNN-based methods like Faster-RCNN [8], YOLOv3 [9], and SSD [10] have achieved state-of-the-art results on some common object detection benchmarks, such as PASCAL VOC [11], MS COCO [12], and so forth.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.