Tracing Englacial Layers in Radargram via Semi-supervised Method: A Preliminary Result

The melting of ice sheets significantly contributes to sea level rise, highlighting the crucial need to comprehend the structure of ice for climate benefits. The stratigraphy of ice sheets revealed through ice layer radargrams gives us a window into historical depth-age correlations and accumulation...

Full description

Bibliographic Details
Published in:Proceedings of the AAAI Symposium Series
Main Authors: Jebeli, Atefeh, Tama, Bayu Adhi, Purushotham, Sanjay, Janeja, Vandana P.
Format: Article in Journal/Newspaper
Language:English
Published: AAAI Press 2024
Subjects:
Online Access:https://ojs.aaai.org/index.php/AAAI-SS/article/view/27653
https://doi.org/10.1609/aaaiss.v2i1.27653
id ftjaaai:oai:ojs.aaai.org:article/27653
record_format openpolar
spelling ftjaaai:oai:ojs.aaai.org:article/27653 2024-02-11T10:04:24+01:00 Tracing Englacial Layers in Radargram via Semi-supervised Method: A Preliminary Result Jebeli, Atefeh Tama, Bayu Adhi Purushotham, Sanjay Janeja, Vandana P. 2024-01-22 application/pdf https://ojs.aaai.org/index.php/AAAI-SS/article/view/27653 https://doi.org/10.1609/aaaiss.v2i1.27653 eng eng AAAI Press https://ojs.aaai.org/index.php/AAAI-SS/article/view/27653/27426 https://ojs.aaai.org/index.php/AAAI-SS/article/view/27653 doi:10.1609/aaaiss.v2i1.27653 Copyright (c) 2023 Association for the Advancement of Artificial Intelligence Proceedings of the AAAI Symposium Series; Vol. 2 No. 1: Proceedings of the 2023 AAAI Fall Symposia; 85-88 2994-4317 Deep Learning Ice Layer Annotation Sea Level Rise Semi-supervised Learning Radargram info:eu-repo/semantics/article info:eu-repo/semantics/publishedVersion Peer-reviewed Article 2024 ftjaaai https://doi.org/10.1609/aaaiss.v2i1.27653 2024-01-27T23:49:56Z The melting of ice sheets significantly contributes to sea level rise, highlighting the crucial need to comprehend the structure of ice for climate benefits. The stratigraphy of ice sheets revealed through ice layer radargrams gives us a window into historical depth-age correlations and accumulation rates. Harnessing this knowledge is not only crucial for interpreting both past and present ice dynamics, especially concerning the Greenland ice sheet, but also for making informed decisions to mitigate the impacts of climate change. Ice layer tracing is prevalently conducted using manual or semi-automatic approaches, requiring significant time and expertise. This study aims to address the need for efficient and precise tracing methods in a two-step process. This is achieved by utilizing an unsupervised annotation method (i.e., ARESELP) to train deep learning models, thereby reducing the need for extensive and time-consuming manual annotations. Four prominent deep learning-based segmentation techniques, namely U-Net, U-Net+VGG19, U-Net+Inception, and Attention U-Net, are benchmarked. Additionally, various thresholding methods such as binary, Otsu, and CLAHE have been explored to achieve optimal enhancement for the true label annotation images. Our preliminary experiments indicate that the combination of attention U-Net with specific processing techniques yields the best performance in terms of the binary IoU metric. Article in Journal/Newspaper Greenland Ice Sheet AAAI Publications (Association for the Advancement of Artificial Intelligence) Greenland Proceedings of the AAAI Symposium Series 2 1 85 88
institution Open Polar
collection AAAI Publications (Association for the Advancement of Artificial Intelligence)
op_collection_id ftjaaai
language English
topic Deep Learning
Ice Layer Annotation
Sea Level Rise
Semi-supervised Learning
Radargram
spellingShingle Deep Learning
Ice Layer Annotation
Sea Level Rise
Semi-supervised Learning
Radargram
Jebeli, Atefeh
Tama, Bayu Adhi
Purushotham, Sanjay
Janeja, Vandana P.
Tracing Englacial Layers in Radargram via Semi-supervised Method: A Preliminary Result
topic_facet Deep Learning
Ice Layer Annotation
Sea Level Rise
Semi-supervised Learning
Radargram
description The melting of ice sheets significantly contributes to sea level rise, highlighting the crucial need to comprehend the structure of ice for climate benefits. The stratigraphy of ice sheets revealed through ice layer radargrams gives us a window into historical depth-age correlations and accumulation rates. Harnessing this knowledge is not only crucial for interpreting both past and present ice dynamics, especially concerning the Greenland ice sheet, but also for making informed decisions to mitigate the impacts of climate change. Ice layer tracing is prevalently conducted using manual or semi-automatic approaches, requiring significant time and expertise. This study aims to address the need for efficient and precise tracing methods in a two-step process. This is achieved by utilizing an unsupervised annotation method (i.e., ARESELP) to train deep learning models, thereby reducing the need for extensive and time-consuming manual annotations. Four prominent deep learning-based segmentation techniques, namely U-Net, U-Net+VGG19, U-Net+Inception, and Attention U-Net, are benchmarked. Additionally, various thresholding methods such as binary, Otsu, and CLAHE have been explored to achieve optimal enhancement for the true label annotation images. Our preliminary experiments indicate that the combination of attention U-Net with specific processing techniques yields the best performance in terms of the binary IoU metric.
format Article in Journal/Newspaper
author Jebeli, Atefeh
Tama, Bayu Adhi
Purushotham, Sanjay
Janeja, Vandana P.
author_facet Jebeli, Atefeh
Tama, Bayu Adhi
Purushotham, Sanjay
Janeja, Vandana P.
author_sort Jebeli, Atefeh
title Tracing Englacial Layers in Radargram via Semi-supervised Method: A Preliminary Result
title_short Tracing Englacial Layers in Radargram via Semi-supervised Method: A Preliminary Result
title_full Tracing Englacial Layers in Radargram via Semi-supervised Method: A Preliminary Result
title_fullStr Tracing Englacial Layers in Radargram via Semi-supervised Method: A Preliminary Result
title_full_unstemmed Tracing Englacial Layers in Radargram via Semi-supervised Method: A Preliminary Result
title_sort tracing englacial layers in radargram via semi-supervised method: a preliminary result
publisher AAAI Press
publishDate 2024
url https://ojs.aaai.org/index.php/AAAI-SS/article/view/27653
https://doi.org/10.1609/aaaiss.v2i1.27653
geographic Greenland
geographic_facet Greenland
genre Greenland
Ice Sheet
genre_facet Greenland
Ice Sheet
op_source Proceedings of the AAAI Symposium Series; Vol. 2 No. 1: Proceedings of the 2023 AAAI Fall Symposia; 85-88
2994-4317
op_relation https://ojs.aaai.org/index.php/AAAI-SS/article/view/27653/27426
https://ojs.aaai.org/index.php/AAAI-SS/article/view/27653
doi:10.1609/aaaiss.v2i1.27653
op_rights Copyright (c) 2023 Association for the Advancement of Artificial Intelligence
op_doi https://doi.org/10.1609/aaaiss.v2i1.27653
container_title Proceedings of the AAAI Symposium Series
container_volume 2
container_issue 1
container_start_page 85
op_container_end_page 88
_version_ 1790600998402654208