FORENSIC ANALYSIS OF DRONE COLLISION WITH TRANSFER LEARNING

Abstract

Drones are one of devices that are used in many different activities. There is a time when drones have accidents, and authorities need to find the cause. Drone forensics is used to determine the cause of an accident. The analysis phase of drone forensics is one of the most important steps in determining accident causes. In this paper, we applied deep learning technique to classify drone collisions. We investigate the use of the InceptionV3 as the deep learning framework. Additionally, this study compares the performance of the proposed method with other techniques, such as MobileNet, VGG, and ResNet, in classifying drone collisions. In this experiment, we also implement transfer learning as well as its fine tuning to speed up the training process and to improve the accuracy value. Additionally, our investigation shows that InceptionV3 outperforms others in terms of accuracy, precision, and F1 scores. [JJCIT 2023; 9(2.000): 175-186]

Keywords

• inception
• mobilenet
• vgg
• restnet
• drone forensics
• transfer learning
• network infrastructure