| Multi-resolution reconstruction of longitudinal streambed footprints using embedded sparse convolutional autoencoders | |
| Yang, Yifan1; Tang, Zihao2; Shao D(邵东)3; Xu, Zhonghou4 | |
| Corresponding Author | Tang, Zihao([email protected]) |
| Source Publication | JOURNAL OF HYDROLOGY
 |
| 2025-06-01 | |
| Volume | 654Pages:18 |
| ISSN | 0022-1694 |
| Abstract | This study introduces an embedded convolutional autoencoder (CAE) architecture designed for the multiresolution reconstruction of longitudinal streambed footprints as sparse heatmaps. Three standalone but interrelated CAEs are trained to achieve double-upsampling, enhancing the heatmaps' spatial resolution and data measurement resolution simultaneously. Transfer learning improves model training efficiency by incorporating a trained model into a larger model at the next level. Cascading the CAEs facilitates a direct pathway for enhancing data quality from the coarsest inputs and recovering fine-grained patterns. Systematic evaluations prove the CAEs' reliability in working individually and collectively. Robustness analyses demonstrate the model's ability to retain field reconstructive quality when subjected to various corrupted inputs, including bulk data loss and spiky noise interference with local measurements at different streambed sections. The model's capacity benefitted from including attention mechanisms (convolutional block attention modules, CBAM) and the adaptive training strategy using crafted loss functions, ensuring efficient extraction and learning of sparse dense patterns and fast reconstruction of physically sound fields. The model architecture's flexibility and scalability are highlighted, proving it suitable for more complex geophysical systems with higher dimensions. The proposed embedded CAE architecture provides a foundational tool for creating digital surrogates of river courses and similar entities, which often involve inherently sparsely distributive data in both spatial and temporal domains. |
| Keyword | Machine learning Streambed footprint Convolutional autoencoder Multi-resolution reconstruction |
| DOI | 10.1016/j.jhydrol.2025.132852 |
| Indexed By | SCI ; EI |
| Language | 英语 |
| WOS ID | WOS:001427259700001 |
| WOS Keyword | RIVER ; PROBABILITY ; MODEL |
| WOS Research Area | Engineering ; Geology ; Water Resources |
| WOS Subject | Engineering, Civil ; Geosciences, Multidisciplinary ; Water Resources |
| Classification | 一类 |
| Ranking | 3 |
| Contributor | Tang, Zihao |
| Citation statistics | |
| Document Type | 期刊论文 |
| Identifier | http://dspace.imech.ac.cn/handle/311007/100159 |
| Collection | 流固耦合系统力学重点实验室 |
| Affiliation | 1.Wuhan Univ, State Key Lab Water Resources Engn & Management, Wuhan, Peoples R China; 2.Univ Auckland, Dept Civil & Environm Engn, Auckland, New Zealand; 3.Chinese Acad Sci, Inst Mech, Key Lab Mech Fluid Solid Coupling Syst, Beijing, Peoples R China; 4.Natl Inst Water & Atmospher Res, Hamilton, New Zealand |
| Recommended Citation GB/T 7714 | Yang, Yifan,Tang, Zihao,Shao D,et al. Multi-resolution reconstruction of longitudinal streambed footprints using embedded sparse convolutional autoencoders[J]. JOURNAL OF HYDROLOGY,2025,654:18.Rp_Au:Tang, Zihao |
| APA | Yang, Yifan,Tang, Zihao,邵东,&Xu, Zhonghou.(2025).Multi-resolution reconstruction of longitudinal streambed footprints using embedded sparse convolutional autoencoders.JOURNAL OF HYDROLOGY,654,18. |
| MLA | Yang, Yifan,et al."Multi-resolution reconstruction of longitudinal streambed footprints using embedded sparse convolutional autoencoders".JOURNAL OF HYDROLOGY 654(2025):18. |
| Files in This Item: | There are no files associated with this item. | |||||
Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.
Edit Comment