Ensemble Deep Learning Using Faster R-CNN and Genetic Algorithm for Vehicle Detection in UAV Images

Abstract

In this paper, an ensemble deep transfer learning (EDTL) based on Faster R-CNN is introduced for the vehicle detection in UAV images. We perform a weighted-averaging ensemble transfer learning comprising six base learners using a ResNet50 that have already pre-trained on ImageNet dataset. The weights of the six base learners as well as the final decision threshold are adaptively optimized via genetic algorithm, to maximize the total accuracy, precision, and recall. Simulation results on AU-AIR dataset demonstrate the superiority of the EDTL against the existing techniques, in terms of the total accuracy, and the trade-off between precision and recall.

KEYWORDS:

• Deep learning
• Ensemble learning
• Faster R-CNN
• Genetic algorithm
• Transfer learning
• Vehicle detection

Additional information

Notes on contributors

Zeinab Ghasemi Darehnaei

Zeinab Ghasemi Darehnaei received her BSc in telecommunication engineering from Semnan University, Faculty of Electrical and Computer Engineering, Semnan, Iran, in 2012. She received her MSc in electronic engineering from Islamic Azad University of Qazvin, Iran, in 2014. She is currently a PhD candidate in electronic engineering at Islamic Azad University of Saveh, Iran. Her research interests include ensemble machine learning, deep learning, image processing, and evolutionary computation. Email: [email protected]

Seyed Mohammad Jalal Rastegar Fatemi

Seyed Mohammad Jalal Rastegar Fatemi was born in Saveh (Iran). He received his MSc in electronic engineering from Islamic Azad University of Arak, Iran, in 2006, and PhD in electronic engineering from Islamic Azad University of Science and Research Branch, Tehran, Iran, in 2011. He joined Islamic Azad University of Saveh in 2006 as a lecturer. His research interests include power electronics and semiconductor device modeling, image processing, deep transfer learning and computer vision.

Seyed Mostafa Mirhassani

Seyed Mostafa Mirhassani received the BEng degree in electronics engineering from Shahrood University of Technology, Shahrood, Iran, and the MEng degree in electronics engineering from Islamic Azad University Science and Research Branch, Tehran, Iran. He received a PhD degree in biomedical engineering in 2015 from the University of Malaya, Kuala Lumpur, Malaysia. Heis currently a faculty member in the Department of Electrical Engineering, Shahrood Branch, Islamic Azad University, Shahrood, Iran. His main areas of research involve speech recognition, medical image processing, soft computing and fuzzy systems. Email: [email protected]

Majid Fouladian

Majid Fouladian received the MS degree from Isfahan University of Technology, Isfahan, Iran in 2008, and the PhD degree from Department of Electrical and Computer Engineering, Islamic Azad University Science and Research Branch, Tehran, Iran in 2016. He is currently a faculty member in the Department of Electrical Engineering, Islamic Azad University Saveh Branch, Saveh, Iran. His research interests include mobile and vehicular communication systems, resource management, and network architectures and protocols. Email: [email protected]