Next Article in Journal
Odour Emissions of Municipal Waste Biogas Plants—Impact of Technological Factors, Air Temperature and Humidity
Next Article in Special Issue
Method for Volume of Irregular Shape Pellets Estimation Using 2D Imaging Measurement
Previous Article in Journal
Students’ Acceptance and Tracking of a New Container-Based Virtual Laboratory
Previous Article in Special Issue
Pedestrian Detection at Night in Infrared Images Using an Attention-Guided Encoder-Decoder Convolutional Neural Network
 
 
Article

Data-Efficient Domain Adaptation for Semantic Segmentation of Aerial Imagery Using Generative Adversarial Networks

1
Robotics and Internet of Things Laboratory, College of Computer and Information Sciences, Prince Sultan University, Riyadh 11586, Saudi Arabia
2
Research Laboratory Smart Electricity & ICT, SEICT, LR18ES44, National Engineering School of Carthage, University of Carthage, Tunis 2035, Tunisia
3
CISTER, INESC-TEC, ISEP, Polytechnic Institute of Porto, 4200-465 Porto, Portugal
*
Author to whom correspondence should be addressed.
Appl. Sci. 2020, 10(3), 1092; https://doi.org/10.3390/app10031092
Received: 17 December 2019 / Revised: 28 January 2020 / Accepted: 29 January 2020 / Published: 6 February 2020
(This article belongs to the Special Issue Applications of Computer Vision in Automation and Robotics)
Despite the significant advances noted in semantic segmentation of aerial imagery, a considerable limitation is blocking its adoption in real cases. If we test a segmentation model on a new area that is not included in its initial training set, accuracy will decrease remarkably. This is caused by the domain shift between the new targeted domain and the source domain used to train the model. In this paper, we addressed this challenge and proposed a new algorithm that uses Generative Adversarial Networks (GAN) architecture to minimize the domain shift and increase the ability of the model to work on new targeted domains. The proposed GAN architecture contains two GAN networks. The first GAN network converts the chosen image from the target domain into a semantic label. The second GAN network converts this generated semantic label into an image that belongs to the source domain but conserves the semantic map of the target image. This resulting image will be used by the semantic segmentation model to generate a better semantic label of the first chosen image. Our algorithm is tested on the ISPRS semantic segmentation dataset and improved the global accuracy by a margin up to 24% when passing from Potsdam domain to Vaihingen domain. This margin can be increased by addition of other labeled data from the target domain. To minimize the cost of supervision in the translation process, we proposed a methodology to use these labeled data efficiently. View Full-Text
Keywords: deep learning; domain adaptation; semantic segmentation; generative adversarial networks; convolutional neural networks; aerial imagery deep learning; domain adaptation; semantic segmentation; generative adversarial networks; convolutional neural networks; aerial imagery
Show Figures

Figure 1

MDPI and ACS Style

Benjdira, B.; Ammar, A.; Koubaa, A.; Ouni, K. Data-Efficient Domain Adaptation for Semantic Segmentation of Aerial Imagery Using Generative Adversarial Networks. Appl. Sci. 2020, 10, 1092. https://doi.org/10.3390/app10031092

AMA Style

Benjdira B, Ammar A, Koubaa A, Ouni K. Data-Efficient Domain Adaptation for Semantic Segmentation of Aerial Imagery Using Generative Adversarial Networks. Applied Sciences. 2020; 10(3):1092. https://doi.org/10.3390/app10031092

Chicago/Turabian Style

Benjdira, Bilel, Adel Ammar, Anis Koubaa, and Kais Ouni. 2020. "Data-Efficient Domain Adaptation for Semantic Segmentation of Aerial Imagery Using Generative Adversarial Networks" Applied Sciences 10, no. 3: 1092. https://doi.org/10.3390/app10031092

Find Other Styles
Note that from the first issue of 2016, MDPI journals use article numbers instead of page numbers. See further details here.

Article Access Map by Country/Region

1
Back to TopTop