AI radar sensor: Creating radar depth sounder images based on generative adversarial network
Significant resources have been spent in collecting and storing large and heterogeneous radar datasets during expensive Arctic and Antarctic fieldwork. The vast majority of data available is unlabeled, and the labeling process is both time-consuming and expensive. One possible alternative to the lab...
Published in: | Sensors |
---|---|
Main Authors: | , , |
Format: | Article in Journal/Newspaper |
Language: | English |
Published: |
MDPI
2019
|
Subjects: | |
Online Access: | https://hdl.handle.net/1969.6/90191 https://doi.org/10.3390/s19245479 |
id |
fttexasamucorpus:oai:tamucc-ir.tdl.org:1969.6/90191 |
---|---|
record_format |
openpolar |
spelling |
fttexasamucorpus:oai:tamucc-ir.tdl.org:1969.6/90191 2023-10-25T01:32:14+02:00 AI radar sensor: Creating radar depth sounder images based on generative adversarial network Rahnemoonfar, Maryam Johnson, Jimmy Paden, John 2019-12-12 application/pdf https://hdl.handle.net/1969.6/90191 https://doi.org/10.3390/s19245479 en_US eng MDPI Rahnemoonfar, M., Johnson, J. and Paden, J., 2019. Ai radar sensor: Creating radar depth sounder images based on generative adversarial network. Sensors, 19(24), p.5479. https://hdl.handle.net/1969.6/90191 https://doi.org/10.3390/s19245479 Attribution 4.0 International http://creativecommons.org/licenses/by/4.0/ convolutional neural networks generative adversarial network ice tracking radar imagery Article 2019 fttexasamucorpus https://doi.org/10.3390/s19245479 2023-09-25T10:19:13Z Significant resources have been spent in collecting and storing large and heterogeneous radar datasets during expensive Arctic and Antarctic fieldwork. The vast majority of data available is unlabeled, and the labeling process is both time-consuming and expensive. One possible alternative to the labeling process is the use of synthetically generated data with artificial intelligence. Instead of labeling real images, we can generate synthetic data based on arbitrary labels. In this way, training data can be quickly augmented with additional images. In this research, we evaluated the performance of synthetically generated radar images based on modified cycle-consistent adversarial networks. We conducted several experiments to test the quality of the generated radar imagery. We also tested the quality of a state-of-the-art contour detection algorithm on synthetic data and different combinations of real and synthetic data. Our experiments show that synthetic radar images generated by generative adversarial network (GAN) can be used in combination with real images for data augmentation and training of deep neural networks. However, the synthetic images generated by GANs cannot be used solely for training a neural network (training on synthetic and testing on real) as they cannot simulate all of the radar characteristics such as noise or Doppler effects. To the best of our knowledge, this is the first work in creating radar sounder imagery based on generative adversarial network. Significant resources have been spent in collecting and storing large and heterogeneous radar datasets during expensive Arctic and Antarctic fieldwork. The vast majority of data available is unlabeled, and the labeling process is both time-consuming and expensive. One possible alternative to the labeling process is the use of synthetically generated data with artificial intelligence. Instead of labeling real images, we can generate synthetic data based on arbitrary labels. In this way, training data can be quickly augmented with additional ... Article in Journal/Newspaper Antarc* Antarctic Arctic Texas A&M University - Corpus Christi: DSpace Repository Arctic Antarctic Sensors 19 24 5479 |
institution |
Open Polar |
collection |
Texas A&M University - Corpus Christi: DSpace Repository |
op_collection_id |
fttexasamucorpus |
language |
English |
topic |
convolutional neural networks generative adversarial network ice tracking radar imagery |
spellingShingle |
convolutional neural networks generative adversarial network ice tracking radar imagery Rahnemoonfar, Maryam Johnson, Jimmy Paden, John AI radar sensor: Creating radar depth sounder images based on generative adversarial network |
topic_facet |
convolutional neural networks generative adversarial network ice tracking radar imagery |
description |
Significant resources have been spent in collecting and storing large and heterogeneous radar datasets during expensive Arctic and Antarctic fieldwork. The vast majority of data available is unlabeled, and the labeling process is both time-consuming and expensive. One possible alternative to the labeling process is the use of synthetically generated data with artificial intelligence. Instead of labeling real images, we can generate synthetic data based on arbitrary labels. In this way, training data can be quickly augmented with additional images. In this research, we evaluated the performance of synthetically generated radar images based on modified cycle-consistent adversarial networks. We conducted several experiments to test the quality of the generated radar imagery. We also tested the quality of a state-of-the-art contour detection algorithm on synthetic data and different combinations of real and synthetic data. Our experiments show that synthetic radar images generated by generative adversarial network (GAN) can be used in combination with real images for data augmentation and training of deep neural networks. However, the synthetic images generated by GANs cannot be used solely for training a neural network (training on synthetic and testing on real) as they cannot simulate all of the radar characteristics such as noise or Doppler effects. To the best of our knowledge, this is the first work in creating radar sounder imagery based on generative adversarial network. Significant resources have been spent in collecting and storing large and heterogeneous radar datasets during expensive Arctic and Antarctic fieldwork. The vast majority of data available is unlabeled, and the labeling process is both time-consuming and expensive. One possible alternative to the labeling process is the use of synthetically generated data with artificial intelligence. Instead of labeling real images, we can generate synthetic data based on arbitrary labels. In this way, training data can be quickly augmented with additional ... |
format |
Article in Journal/Newspaper |
author |
Rahnemoonfar, Maryam Johnson, Jimmy Paden, John |
author_facet |
Rahnemoonfar, Maryam Johnson, Jimmy Paden, John |
author_sort |
Rahnemoonfar, Maryam |
title |
AI radar sensor: Creating radar depth sounder images based on generative adversarial network |
title_short |
AI radar sensor: Creating radar depth sounder images based on generative adversarial network |
title_full |
AI radar sensor: Creating radar depth sounder images based on generative adversarial network |
title_fullStr |
AI radar sensor: Creating radar depth sounder images based on generative adversarial network |
title_full_unstemmed |
AI radar sensor: Creating radar depth sounder images based on generative adversarial network |
title_sort |
ai radar sensor: creating radar depth sounder images based on generative adversarial network |
publisher |
MDPI |
publishDate |
2019 |
url |
https://hdl.handle.net/1969.6/90191 https://doi.org/10.3390/s19245479 |
geographic |
Arctic Antarctic |
geographic_facet |
Arctic Antarctic |
genre |
Antarc* Antarctic Arctic |
genre_facet |
Antarc* Antarctic Arctic |
op_relation |
Rahnemoonfar, M., Johnson, J. and Paden, J., 2019. Ai radar sensor: Creating radar depth sounder images based on generative adversarial network. Sensors, 19(24), p.5479. https://hdl.handle.net/1969.6/90191 https://doi.org/10.3390/s19245479 |
op_rights |
Attribution 4.0 International http://creativecommons.org/licenses/by/4.0/ |
op_doi |
https://doi.org/10.3390/s19245479 |
container_title |
Sensors |
container_volume |
19 |
container_issue |
24 |
container_start_page |
5479 |
_version_ |
1780727840387366912 |