Table of Content

Deep Learning for Marine and Underwater Environment: Theory, Method, and Applications

Submission Deadline: 31 October 2023 Submit to Special Issue

Guest Editors

Prof. Liguo Zhang, Harbin Engineering University, China.
Prof. Shengke Wang, Ocean University of China, China.
Prof. Sizhao Li, The Chinese University of Hong Kong, Hong Kong, China.


Marine science is an interdisciplinary field at the intersection of physics, chemistry, biology, geography, mathematics, computer science, information engineering, and mechatronik. In recent years, deep learning technology (DLT) developed in this field is widely used in ocean exploration and development, mainly in resource exploration, environmental protection and naval construction. Although the current research and development methods for ships, structures and vehicles are relatively abundant, many tasks, such as deep-sea resources modeling, underwater 3D reconstruction, marine sensor network modeling, sonar signal processing, marine environmental analysis, are limited by the machine learning algorithm and the computing power of computer. And the experimental methods are closely relevant to the computation cost and speed. With the rapid development of computer hardware and artificial intelligence technology, how to combine DLT with research topics related to the marine field and the underwater environment become a most challenging area of research directions in marine science and engineering. Therefore, novel multidimensional, multiscale, and multimodal system or scene modeling methods are being continuously developed to describe these complex tasks.


This Special Issue focuses on the recent advances and challenges for DL methods and applications of various tasks in marine and underwater environment, with the goal of realizing efficient and diversified research and development of ships, structures and marine environment. Contributions may present and solve open technical problems and integrate novel solutions efficiently.


Topics of interest include but are not restricted to:

- Marine 3D scene modeling using DLT

- Marine sensor network modeling using DLT

- Hydrodynamics modeling and simulation calculation based on DLT

- Ship structure modeling and simulation calculation based on DLT

- Underwater computer vision and image processing using DLT

- Underwater acoustical signal processing using DLT

- Underwater multimodal data processing using DLT

- Design and application of sea UAV and UUV based on DLT

- The other related topics

Published Papers

  • Open Access


    PanopticUAV: Panoptic Segmentation of UAV Images for Marine Environment Monitoring

    Yuling Dou, Fengqin Yao, Xiandong Wang, Liang Qu, Long Chen, Zhiwei Xu, Laihui Ding, Leon Bevan Bullock, Guoqiang Zhong, Shengke Wang
    CMES-Computer Modeling in Engineering & Sciences, Vol.138, No.1, pp. 1001-1014, 2024, DOI:10.32604/cmes.2023.027764
    (This article belongs to this Special Issue: Deep Learning for Marine and Underwater Environment: Theory, Method, and Applications)
    Abstract UAV marine monitoring plays an essential role in marine environmental protection because of its flexibility and convenience, low cost and convenient maintenance. In marine environmental monitoring, the similarity between objects such as oil spill and sea surface, Spartina alterniflora and algae is high, and the effect of the general segmentation algorithm is poor, which brings new challenges to the segmentation of UAV marine images. Panoramic segmentation can do object detection and semantic segmentation at the same time, which can well solve the polymorphism problem of objects in UAV ocean images. Currently, there are few studies on UAV marine image recognition… More >

  • Open Access


    Filter Bank Networks for Few-Shot Class-Incremental Learning

    Yanzhao Zhou, Binghao Liu, Yiran Liu, Jianbin Jiao
    CMES-Computer Modeling in Engineering & Sciences, Vol.137, No.1, pp. 647-668, 2023, DOI:10.32604/cmes.2023.026745
    (This article belongs to this Special Issue: Deep Learning for Marine and Underwater Environment: Theory, Method, and Applications)
    Abstract Deep Convolution Neural Networks (DCNNs) can capture discriminative features from large datasets. However, how to incrementally learn new samples without forgetting old ones and recognize novel classes that arise in the dynamically changing world, e.g., classifying newly discovered fish species, remains an open problem. We address an even more challenging and realistic setting of this problem where new class samples are insufficient, i.e., Few-Shot Class-Incremental Learning (FSCIL). Current FSCIL methods augment the training data to alleviate the overfitting of novel classes. By contrast, we propose Filter Bank Networks (FBNs) that augment the learnable filters to capture fine-detailed features for adapting… More >

Share Link