Joint Learning of Underwater Terrain Matching and Suitability Analysis via SO(2) Elevation Transformer
Underwater navigation reliability heavily depends on accurate terrain evaluation. While conventional approaches treat terrain matching and suitability assessment as separate processes, this separation often leads to inconsistent and imprecise terrain evaluations. To address these limitations, we pro...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10902406/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850028739252977664 |
|---|---|
| author | Yan Han Gang Fan Qichen Yan Pengyun Chen Xiaolong Yan Tinghai Yan Guoguang Chen |
| author_facet | Yan Han Gang Fan Qichen Yan Pengyun Chen Xiaolong Yan Tinghai Yan Guoguang Chen |
| author_sort | Yan Han |
| collection | DOAJ |
| description | Underwater navigation reliability heavily depends on accurate terrain evaluation. While conventional approaches treat terrain matching and suitability assessment as separate processes, this separation often leads to inconsistent and imprecise terrain evaluations. To address these limitations, we propose an integrated framework that employs a unified terrain map encoder to simultaneously handle both matching and suitability analysis tasks. At the core of our framework lies the SO(2) Elevation Embedding Transformer (SEET), which combines rotation-equivariant CNN with elevation embeddings. The SEET encoder is pre-trained through self-supervised contrastive learning on underwater elevation data, eliminating the need for manual labeling. Our extensive experimental validation demonstrates the framework’s effectiveness, showing superior matching accuracy with minimal navigation deviation. The distinct performance gap observed between suitable and unsuitable regions further validates the effectiveness of our suitability analysis approach. |
| format | Article |
| id | doaj-art-9e68e57b00f543d0b49e4ce06e706211 |
| institution | DOAJ |
| issn | 2169-3536 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-9e68e57b00f543d0b49e4ce06e7062112025-08-20T02:59:45ZengIEEEIEEE Access2169-35362025-01-0113383013831610.1109/ACCESS.2025.354556310902406Joint Learning of Underwater Terrain Matching and Suitability Analysis via SO(2) Elevation TransformerYan Han0https://orcid.org/0000-0002-1660-0065Gang Fan1Qichen Yan2https://orcid.org/0009-0008-0739-7537Pengyun Chen3https://orcid.org/0000-0003-2157-281XXiaolong Yan4https://orcid.org/0000-0002-0395-8973Tinghai Yan5Guoguang Chen6https://orcid.org/0000-0002-2319-8852Department of Mechatronic Engineering, North University of China, Taiyuan, ChinaDepartment of Electrical Engineering, Taiyuan Institute of Technology, Taiyuan, ChinaDepartment of Mechatronic Engineering, North University of China, Taiyuan, ChinaDepartment of Mechatronic Engineering, North University of China, Taiyuan, ChinaDepartment of Mechatronic Engineering, North University of China, Taiyuan, ChinaQiyuan Laboratory, Beijing, ChinaDepartment of Mechatronic Engineering, North University of China, Taiyuan, ChinaUnderwater navigation reliability heavily depends on accurate terrain evaluation. While conventional approaches treat terrain matching and suitability assessment as separate processes, this separation often leads to inconsistent and imprecise terrain evaluations. To address these limitations, we propose an integrated framework that employs a unified terrain map encoder to simultaneously handle both matching and suitability analysis tasks. At the core of our framework lies the SO(2) Elevation Embedding Transformer (SEET), which combines rotation-equivariant CNN with elevation embeddings. The SEET encoder is pre-trained through self-supervised contrastive learning on underwater elevation data, eliminating the need for manual labeling. Our extensive experimental validation demonstrates the framework’s effectiveness, showing superior matching accuracy with minimal navigation deviation. The distinct performance gap observed between suitable and unsuitable regions further validates the effectiveness of our suitability analysis approach.https://ieeexplore.ieee.org/document/10902406/Rotation invariancesuitability analysisSO(2) transformerterrain evaluationunderwater navigation |
| spellingShingle | Yan Han Gang Fan Qichen Yan Pengyun Chen Xiaolong Yan Tinghai Yan Guoguang Chen Joint Learning of Underwater Terrain Matching and Suitability Analysis via SO(2) Elevation Transformer IEEE Access Rotation invariance suitability analysis SO(2) transformer terrain evaluation underwater navigation |
| title | Joint Learning of Underwater Terrain Matching and Suitability Analysis via SO(2) Elevation Transformer |
| title_full | Joint Learning of Underwater Terrain Matching and Suitability Analysis via SO(2) Elevation Transformer |
| title_fullStr | Joint Learning of Underwater Terrain Matching and Suitability Analysis via SO(2) Elevation Transformer |
| title_full_unstemmed | Joint Learning of Underwater Terrain Matching and Suitability Analysis via SO(2) Elevation Transformer |
| title_short | Joint Learning of Underwater Terrain Matching and Suitability Analysis via SO(2) Elevation Transformer |
| title_sort | joint learning of underwater terrain matching and suitability analysis via so 2 elevation transformer |
| topic | Rotation invariance suitability analysis SO(2) transformer terrain evaluation underwater navigation |
| url | https://ieeexplore.ieee.org/document/10902406/ |
| work_keys_str_mv | AT yanhan jointlearningofunderwaterterrainmatchingandsuitabilityanalysisviaso2elevationtransformer AT gangfan jointlearningofunderwaterterrainmatchingandsuitabilityanalysisviaso2elevationtransformer AT qichenyan jointlearningofunderwaterterrainmatchingandsuitabilityanalysisviaso2elevationtransformer AT pengyunchen jointlearningofunderwaterterrainmatchingandsuitabilityanalysisviaso2elevationtransformer AT xiaolongyan jointlearningofunderwaterterrainmatchingandsuitabilityanalysisviaso2elevationtransformer AT tinghaiyan jointlearningofunderwaterterrainmatchingandsuitabilityanalysisviaso2elevationtransformer AT guoguangchen jointlearningofunderwaterterrainmatchingandsuitabilityanalysisviaso2elevationtransformer |