TransSounder: A Hybrid TransUNet-TransFuse Architectural Framework for Semantic Segmentation of Radar Sounder Data
Radar sounders (RSs) are nadir-looking sensors operating in high frequency (HF) or very high frequency (VHF) bands that profile subsurface targets to retrieve miscellaneous scientific information. Due to the complex electromagnetic interaction between backscattered returns, the interpretation of RS...
Published in: | IEEE Transactions on Geoscience and Remote Sensing |
---|---|
Main Authors: | , |
Other Authors: | , |
Format: | Article in Journal/Newspaper |
Language: | English |
Published: |
country:USA
2022
|
Subjects: | |
Online Access: | https://hdl.handle.net/11572/354883 https://doi.org/10.1109/TGRS.2022.3180761 https://ieeexplore.ieee.org/document/9791273/authors#authors |
id |
ftutrentoiris:oai:iris.unitn.it:11572/354883 |
---|---|
record_format |
openpolar |
spelling |
ftutrentoiris:oai:iris.unitn.it:11572/354883 2024-02-11T10:04:53+01:00 TransSounder: A Hybrid TransUNet-TransFuse Architectural Framework for Semantic Segmentation of Radar Sounder Data Ghosh R. Bovolo F. Ghosh, R. Bovolo, F. 2022 STAMPA https://hdl.handle.net/11572/354883 https://doi.org/10.1109/TGRS.2022.3180761 https://ieeexplore.ieee.org/document/9791273/authors#authors eng eng country:USA info:eu-repo/semantics/altIdentifier/wos/WOS:000814744100007 volume:60 firstpage:1 lastpage:13 numberofpages:13 journal:IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING https://hdl.handle.net/11572/354883 doi:10.1109/TGRS.2022.3180761 info:eu-repo/semantics/altIdentifier/scopus/2-s2.0-85131737903 https://ieeexplore.ieee.org/document/9791273/authors#authors info:eu-repo/semantics/openAccess Multichannel Coherent Radar Depth Sounder (MCoRDS) radar sounder (RS) semantic segmentation sequence-to-sequence model transformers TransFuse TransUNet info:eu-repo/semantics/article 2022 ftutrentoiris https://doi.org/10.1109/TGRS.2022.3180761 2024-01-23T23:06:10Z Radar sounders (RSs) are nadir-looking sensors operating in high frequency (HF) or very high frequency (VHF) bands that profile subsurface targets to retrieve miscellaneous scientific information. Due to the complex electromagnetic interaction between backscattered returns, the interpretation of RS data is challenging. The investigations of ice-sheet subsurface structures require automatic techniques to account for both the sequential spatial distribution of subsurface targets and relevant statistical properties embedded in RS signals. Automatic techniques exist for characterizing these targets either related to probabilistic inference models or convolutional neural network (CNN) deep learning methods. Unfortunately, CNN-based methods capture local spatial context and merely model the global spatial context. In contrast to CNN, the transformer-based models are reliable architectures for capturing long-range sequence-to-sequence global spatial contextual prior. Motivated by the aforementioned fact, we propose a novel transformer-based semantic segmentation architecture named TransSounder to effectively encode the sequential structures of the RS signals. The TransSounder was constructed on a hybrid TransUNet-TransFuse architectural framework to systematically augment the modules from TransUNet and TransFuse architectures. Experimental results obtained using the Multichannel Coherent Radar Depth Sounder (MCoRDS) dataset confirms the robustness and capability of transformers to accurately characterize the different subsurface targets. Article in Journal/Newspaper Ice Sheet Università degli Studi di Trento: CINECA IRIS IEEE Transactions on Geoscience and Remote Sensing 60 1 13 |
institution |
Open Polar |
collection |
Università degli Studi di Trento: CINECA IRIS |
op_collection_id |
ftutrentoiris |
language |
English |
topic |
Multichannel Coherent Radar Depth Sounder (MCoRDS) radar sounder (RS) semantic segmentation sequence-to-sequence model transformers TransFuse TransUNet |
spellingShingle |
Multichannel Coherent Radar Depth Sounder (MCoRDS) radar sounder (RS) semantic segmentation sequence-to-sequence model transformers TransFuse TransUNet Ghosh R. Bovolo F. TransSounder: A Hybrid TransUNet-TransFuse Architectural Framework for Semantic Segmentation of Radar Sounder Data |
topic_facet |
Multichannel Coherent Radar Depth Sounder (MCoRDS) radar sounder (RS) semantic segmentation sequence-to-sequence model transformers TransFuse TransUNet |
description |
Radar sounders (RSs) are nadir-looking sensors operating in high frequency (HF) or very high frequency (VHF) bands that profile subsurface targets to retrieve miscellaneous scientific information. Due to the complex electromagnetic interaction between backscattered returns, the interpretation of RS data is challenging. The investigations of ice-sheet subsurface structures require automatic techniques to account for both the sequential spatial distribution of subsurface targets and relevant statistical properties embedded in RS signals. Automatic techniques exist for characterizing these targets either related to probabilistic inference models or convolutional neural network (CNN) deep learning methods. Unfortunately, CNN-based methods capture local spatial context and merely model the global spatial context. In contrast to CNN, the transformer-based models are reliable architectures for capturing long-range sequence-to-sequence global spatial contextual prior. Motivated by the aforementioned fact, we propose a novel transformer-based semantic segmentation architecture named TransSounder to effectively encode the sequential structures of the RS signals. The TransSounder was constructed on a hybrid TransUNet-TransFuse architectural framework to systematically augment the modules from TransUNet and TransFuse architectures. Experimental results obtained using the Multichannel Coherent Radar Depth Sounder (MCoRDS) dataset confirms the robustness and capability of transformers to accurately characterize the different subsurface targets. |
author2 |
Ghosh, R. Bovolo, F. |
format |
Article in Journal/Newspaper |
author |
Ghosh R. Bovolo F. |
author_facet |
Ghosh R. Bovolo F. |
author_sort |
Ghosh R. |
title |
TransSounder: A Hybrid TransUNet-TransFuse Architectural Framework for Semantic Segmentation of Radar Sounder Data |
title_short |
TransSounder: A Hybrid TransUNet-TransFuse Architectural Framework for Semantic Segmentation of Radar Sounder Data |
title_full |
TransSounder: A Hybrid TransUNet-TransFuse Architectural Framework for Semantic Segmentation of Radar Sounder Data |
title_fullStr |
TransSounder: A Hybrid TransUNet-TransFuse Architectural Framework for Semantic Segmentation of Radar Sounder Data |
title_full_unstemmed |
TransSounder: A Hybrid TransUNet-TransFuse Architectural Framework for Semantic Segmentation of Radar Sounder Data |
title_sort |
transsounder: a hybrid transunet-transfuse architectural framework for semantic segmentation of radar sounder data |
publisher |
country:USA |
publishDate |
2022 |
url |
https://hdl.handle.net/11572/354883 https://doi.org/10.1109/TGRS.2022.3180761 https://ieeexplore.ieee.org/document/9791273/authors#authors |
genre |
Ice Sheet |
genre_facet |
Ice Sheet |
op_relation |
info:eu-repo/semantics/altIdentifier/wos/WOS:000814744100007 volume:60 firstpage:1 lastpage:13 numberofpages:13 journal:IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING https://hdl.handle.net/11572/354883 doi:10.1109/TGRS.2022.3180761 info:eu-repo/semantics/altIdentifier/scopus/2-s2.0-85131737903 https://ieeexplore.ieee.org/document/9791273/authors#authors |
op_rights |
info:eu-repo/semantics/openAccess |
op_doi |
https://doi.org/10.1109/TGRS.2022.3180761 |
container_title |
IEEE Transactions on Geoscience and Remote Sensing |
container_volume |
60 |
container_start_page |
1 |
op_container_end_page |
13 |
_version_ |
1790601638186057728 |