Data Augmentation Analysis in Vehicle Detection from Aerial Videos

Author(s)

• Quynh M. Chung, Vietnam National UniversityFollow
• Tien D. Le, Vietnam National UniversityFollow
• Thin V. Dang, Vietnam National UniversityFollow
• Nguyen D. Vo, Vietnam National UniversityFollow
• Khang Nguyen, Vietnam National UniversityFollow
• Tam V. Nguyen, University of DaytonFollow

Document Type

Conference Paper

Publication Date

7-15-2020

Publication Source

2020 RIVF International Conference on Computing and Communication Technologies (RIVF)

Abstract

Recent growth in deep learning and computer vision has opened up many opportunities for advanced intelligent systems. While the dataset quality plays a crucial role in the training phase and also affects the performance of a model, creating a reliable and diverse dataset appears to be challenging. Therefore, data augmentation can be used as a preprocessing step with a view to tackling the problem. In this paper, we conduct a comprehensive analysis on different augmentation strategies to investigate their impact on vehicle detection in top-down videos captured by drones. Our experiments show that by adding similar data, randomly rotating and cropping images with respect to the model's input size, we have remarkably increased the accuracy of YOLOv3, one of the state-of-the-art and real-time object detection methods.

ISBN/ISSN

2162-786X

Comments

The article available for download is the authors' accepted manuscript, provided in compliance with the publisher's policy on self-archiving. Permission documentation is on file.

Publisher

IEEE

Keywords

drone, real-time object detection, vehicle detection, YOLO, data augmentation

eCommons Citation

Chung, Quynh M.; Le, Tien D.; Dang, Thin V.; Vo, Nguyen D.; Nguyen, Khang; and Nguyen, Tam V., "Data Augmentation Analysis in Vehicle Detection from Aerial Videos" (2020). Computer Science Faculty Publications. 209.
https://ecommons.udayton.edu/cps_fac_pub/209