A Hybrid Dropout Method for High-Precision Seafloor Topography Reconstruction and Uncertainty Quantification
Seafloor topography super-resolution reconstruction is critical for marine resource exploration, geological monitoring, and navigation safety. However, sparse acoustic data frequently result in the loss of high-frequency details, and traditional deep learning models exhibit limitations in uncertaint...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-05-01
|
| Series: | Applied Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3417/15/11/6113 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850160619080122368 |
|---|---|
| author | Xinye Cui Houpu Li Yanting Yu Shaofeng Bian Guojun Zhai |
| author_facet | Xinye Cui Houpu Li Yanting Yu Shaofeng Bian Guojun Zhai |
| author_sort | Xinye Cui |
| collection | DOAJ |
| description | Seafloor topography super-resolution reconstruction is critical for marine resource exploration, geological monitoring, and navigation safety. However, sparse acoustic data frequently result in the loss of high-frequency details, and traditional deep learning models exhibit limitations in uncertainty quantification, impeding their practical application. To address these challenges, this study systematically investigates the combined effects of various regularization strategies and uncertainty quantification modules. It proposes a hybrid dropout model that jointly optimizes high-precision reconstruction and uncertainty estimation. The model integrates residual blocks, squeeze-and-excitation (SE) modules, and a multi-scale feature extraction network while employing Monte Carlo Dropout (MC-Dropout) alongside heteroscedastic noise modeling to dynamically gate the uncertainty quantification process. By adaptively modulating the regularization strength based on feature activations, the model preserves high-frequency information and accurately estimates predictive uncertainty. The experimental results demonstrate significant improvements in the Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Peak Signal-to-Noise Ratio (PSNR). Compared to conventional dropout architectures, the proposed method achieves a PSNR increase of 46.5% to 60.5% in test regions with a marked reduction in artifacts. Overall, the synergistic effect of employed regularization strategies and uncertainty quantification modules substantially enhances detail recovery and robustness in complex seafloor topography reconstruction, offering valuable theoretical insights and practical guidance for further optimization of deep learning models in challenging applications. |
| format | Article |
| id | doaj-art-e22c8e6f11cb4829a22d3f2da90f10c3 |
| institution | OA Journals |
| issn | 2076-3417 |
| language | English |
| publishDate | 2025-05-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Applied Sciences |
| spelling | doaj-art-e22c8e6f11cb4829a22d3f2da90f10c32025-08-20T02:23:06ZengMDPI AGApplied Sciences2076-34172025-05-011511611310.3390/app15116113A Hybrid Dropout Method for High-Precision Seafloor Topography Reconstruction and Uncertainty QuantificationXinye Cui0Houpu Li1Yanting Yu2Shaofeng Bian3Guojun Zhai4School of Electrical Engineering, Naval University of Engineering, Wuhan 430033, ChinaSchool of Electrical Engineering, Naval University of Engineering, Wuhan 430033, ChinaSchool of Electrical Engineering, Naval University of Engineering, Wuhan 430033, ChinaSchool of Electrical Engineering, Naval University of Engineering, Wuhan 430033, ChinaKey Laboratory of Geological Exploration and Evaluation, Ministry of Education, China University of Geosciences, Wuhan 430074, ChinaSeafloor topography super-resolution reconstruction is critical for marine resource exploration, geological monitoring, and navigation safety. However, sparse acoustic data frequently result in the loss of high-frequency details, and traditional deep learning models exhibit limitations in uncertainty quantification, impeding their practical application. To address these challenges, this study systematically investigates the combined effects of various regularization strategies and uncertainty quantification modules. It proposes a hybrid dropout model that jointly optimizes high-precision reconstruction and uncertainty estimation. The model integrates residual blocks, squeeze-and-excitation (SE) modules, and a multi-scale feature extraction network while employing Monte Carlo Dropout (MC-Dropout) alongside heteroscedastic noise modeling to dynamically gate the uncertainty quantification process. By adaptively modulating the regularization strength based on feature activations, the model preserves high-frequency information and accurately estimates predictive uncertainty. The experimental results demonstrate significant improvements in the Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Peak Signal-to-Noise Ratio (PSNR). Compared to conventional dropout architectures, the proposed method achieves a PSNR increase of 46.5% to 60.5% in test regions with a marked reduction in artifacts. Overall, the synergistic effect of employed regularization strategies and uncertainty quantification modules substantially enhances detail recovery and robustness in complex seafloor topography reconstruction, offering valuable theoretical insights and practical guidance for further optimization of deep learning models in challenging applications.https://www.mdpi.com/2076-3417/15/11/6113super-resolution reconstructionuncertainty quantificationseafloor topographydeep learningheteroscedastic noise modeling |
| spellingShingle | Xinye Cui Houpu Li Yanting Yu Shaofeng Bian Guojun Zhai A Hybrid Dropout Method for High-Precision Seafloor Topography Reconstruction and Uncertainty Quantification Applied Sciences super-resolution reconstruction uncertainty quantification seafloor topography deep learning heteroscedastic noise modeling |
| title | A Hybrid Dropout Method for High-Precision Seafloor Topography Reconstruction and Uncertainty Quantification |
| title_full | A Hybrid Dropout Method for High-Precision Seafloor Topography Reconstruction and Uncertainty Quantification |
| title_fullStr | A Hybrid Dropout Method for High-Precision Seafloor Topography Reconstruction and Uncertainty Quantification |
| title_full_unstemmed | A Hybrid Dropout Method for High-Precision Seafloor Topography Reconstruction and Uncertainty Quantification |
| title_short | A Hybrid Dropout Method for High-Precision Seafloor Topography Reconstruction and Uncertainty Quantification |
| title_sort | hybrid dropout method for high precision seafloor topography reconstruction and uncertainty quantification |
| topic | super-resolution reconstruction uncertainty quantification seafloor topography deep learning heteroscedastic noise modeling |
| url | https://www.mdpi.com/2076-3417/15/11/6113 |
| work_keys_str_mv | AT xinyecui ahybriddropoutmethodforhighprecisionseafloortopographyreconstructionanduncertaintyquantification AT houpuli ahybriddropoutmethodforhighprecisionseafloortopographyreconstructionanduncertaintyquantification AT yantingyu ahybriddropoutmethodforhighprecisionseafloortopographyreconstructionanduncertaintyquantification AT shaofengbian ahybriddropoutmethodforhighprecisionseafloortopographyreconstructionanduncertaintyquantification AT guojunzhai ahybriddropoutmethodforhighprecisionseafloortopographyreconstructionanduncertaintyquantification AT xinyecui hybriddropoutmethodforhighprecisionseafloortopographyreconstructionanduncertaintyquantification AT houpuli hybriddropoutmethodforhighprecisionseafloortopographyreconstructionanduncertaintyquantification AT yantingyu hybriddropoutmethodforhighprecisionseafloortopographyreconstructionanduncertaintyquantification AT shaofengbian hybriddropoutmethodforhighprecisionseafloortopographyreconstructionanduncertaintyquantification AT guojunzhai hybriddropoutmethodforhighprecisionseafloortopographyreconstructionanduncertaintyquantification |