Recognizing military vehicles in social media images using deep learning

Show full item record

Permalink

http://hdl.handle.net/10138/211810

Citation

Hiippala , T 2017 , Recognizing military vehicles in social media images using deep learning . in Proceedings of the 2017 IEEE International Conference on Intelligence and Security Informatics . IEEE , Piscataway, NJ , pp. 60-65 , IEEE International Conference on Intelligence and Security Informatics , Beijing , China , 22/07/2017 . DOI: 10.1109/ISI.2017.8004875

Title: Recognizing military vehicles in social media images using deep learning
Author: Hiippala, Tuomo
Contributor: University of Helsinki, Department of Geosciences and Geography
Belongs to series: Proceedings of the 2017 IEEE International Conference on Intelligence and Security Informatics
ISBN: 978-1-5090-6728-2
978-1-5090-6727-5
Abstract: This paper presents a system that uses machine learning to recognize military vehicles in social media images. To do so, the system draws on recent advances in applying deep neural networks to computer vision tasks, while also making extensive use of openly available libraries, models and data. Training a vehicle recognition system over three classes, the paper reports on two experiments that use different architectures and strategies to overcome the challenges of working with limited training data: data augmentation and transfer learning. The results show that transfer learning outperforms data augmentation, achieving an average accuracy of 95.18% using 10-fold cross-validation, while also generalizing well on a separate testing set consisting of social media content.
URI: http://hdl.handle.net/10138/211810
Date: 2017
Subject: 113 Computer and information sciences
computer vision
machine learning
deep learning
open source intelligence
neural network
social media
518 Media and communications
social media
Rights:


Files in this item

Total number of downloads: Loading...

Files Size Format View
hiippala2017_isi.pdf 2.999Mb PDF View/Open

This item appears in the following Collection(s)

Show full item record