Sonar Image Datasets: A Comprehensive Survey of Resources, Challenges, and Applications
Instituto de Ciências Exatas e Tecnologia (ICET). Universidade Federal do Amazonas(巴西亚马逊联邦大学理学院(ICET))
Larissa S. Gomes, Gustavo P. Almeida, Bryan U. Moreira, Marco Quiroz, Breno Xavier, Lucas Soares, Stephanie L. Brião, Felipe G. Oliveira, Paulo L. J. Drews-Jr
备注:
Published in the Conference on Graphics, Patterns and Images (SIBGRAPI). This 4-page paper presents a timeline of publicly available datasets up to the year 2025
Sonar images are relevant for advancing underwater exploration, autonomous navigation, and ecosystem monitoring. However, the progress depends on data availability. The scarcity of publicly available, well-annotated sonar image datasets creates a significant bottleneck for the development of robust machine learning models. This paper presents a comprehensive and concise review of the current landscape of sonar image datasets, seeking not only to catalog existing resources but also to contextualize them, identify gaps, and provide a clear roadmap, serving as a base guide for researchers of any kind who wish to start or advance in the field of underwater acoustic data analysis. We mapped publicly accessible datasets across various sonar modalities, including Side Scan Sonar (SSS), Forward-Looking Sonar (FLS), Synthetic Aperture Sonar (SAS), Multibeam Echo Sounder (MBES), and Dual-Frequency Identification Sonar (DIDSON). An analysis was conducted on applications such as classification, detection, segmentation, and 3D reconstruction. This work focuses on state-of-the-art advancements, incorporating newly released datasets. The findings are synthesized into a master table and a chronological timeline, offering a clear and accessible comparison of characteristics, sizes, and annotation details datasets.
China Unicom (Shanghai) Industry Internet Co., Ltd.(中国联通(上海)产业互联网有限公司);College of Intelligent Systems Science and Engineering, Harbin Engineering University(哈尔滨工程大学智能系统科学与工程学院);Shanghai Graphic Design Information Co., Ltd(上海平面设计信息有限公司)
本文填补了Segment Anything Model (SAM)在声纳成像领域应用的研究空白,通过多种设置评估其性能,并采用带提示和语义分割的微调方法显著提升了其在自动化分割任务中的表现。
Sonar-based Deep Learning in Underwater Robotics: Overview, Robustness and Challenges
OceanScan Marine Systems & Technology(OceanScan海洋系统与技术公司);INESC INOV-Lab(INESC INOV实验室);ISRC (ISEP/P.PORTO)(ISRC(波尔图高等工程学院/波尔图大学));Faculty of Engineering, University of Porto (FEUP)(波尔图大学工程学院(FEUP))
Can Optical Denoising Clean Sonar Images? A Benchmark and Fusion Approach
Hangzhou Institute of Technology, Xidian University(杭州理工大学);Department of Engineering, Shenzhen MSU-BIT University(深圳MSU-BIT大学工程系);School of Information Engineering, Jiangxi University of Science(江西理工大学信息工程学院);Institute of Deep-sea Science(深海科学研究院);School of Computer Science, Northwestern Polytechnical University(西北工业大学计算机学院);College of Cyber Security, Jinan University(广州大学网络安全学院)
Deep Neural Networks for Marine Debris Detection in Sonar Images
Heriot-Watt University(赫瑞-瓦特大学)
垃圾和废物处理是人类目前面临的最大挑战之一。适当的废物处理和回收利用是任何可持续发展社区的必要条件,在许多沿海地区,漂浮或淹没的垃圾造成了严重的水污染。这就是所谓的海洋垃圾。淹没的海洋废弃物威胁海洋生物,对于浅水沿海地区,它还可能威胁渔船[Iñiguez et al. 2016,Renewable and Sustainable Energy Reviews]。淹没的海洋废弃物通常会在环境中停留很长时间(20年以上),并且由可以回收的材料组成,例如金属,塑料,玻璃等。本文对DNN在FLS图像海洋垃圾检测中的应用进行了综合评价,并对相关问题如图像分类、匹配和检测建议等进行了研究。我们这样做的2069 FLS图像的数据集,我们捕获的ARIS Explorer 3000传感器上的海洋垃圾物体躺在地板上的一个小水箱。我们用于生成此数据集的对象包含典型的家庭海洋垃圾和干扰海洋对象(轮胎,钩子,阀门等),分为10个类和一个背景类。我们的研究结果表明,对于评估的任务,DNN是一种优于相应的最先进技术的技术。特别是对于匹配和检测建议任务,DNN有很大的收益。我们还研究了样本复杂性和对象大小在许多任务中的影响,这对从业者来说是有价值的信息。我们希望我们的研究结果将推进使用自主水下航行器自动调查,检测和收集水下环境中的海洋垃圾的目标。
FTibSuite: A Comprehensive Resource Suite for Tibetan Vision-Language Modeling
Hainan International College, Minzu University of China(民族大学海南国际学院);School of Information Engineering, Minzu University of China(民族大学信息工程学院);Shanghai Jiao Tong University(上海交通大学);Institute of Automation, Chinese Academy of Sciences(中国科学院自动化研究所)
Expanding on the BRIAR Dataset: A Comprehensive Whole Body Biometric Recognition Resource at Extreme Distances and Real-World Scenarios (Collections 1-4)
Oak Ridge National Laboratory(橡树岭国家实验室);IARPA(美国情报高级研究计划局)
Generative and reproducible benchmarks for comprehensive evaluation of machine learning classifiers
Institute for Biomedical Informatics, University of Pennsylvania(宾夕法尼亚大学生物医学信息学研究所);AGH University of Science and Technology(AGH科技大学)
了解机器学习(ML)算法的优点和缺点对于确定其应用范围至关重要。在这里,我们介绍了DIverse和GENerative ML Benchmark(DIGEN)-一个综合数据集的集合,用于对二进制结果分类的机器学习算法进行全面,可重复和可解释的基准测试。DIGEN资源由40个数学函数组成,这些函数将连续特征映射到离散端点,以创建合成数据集。这40个函数是使用启发式算法发现的,该算法旨在最大限度地提高多个流行的机器学习算法之间的性能差异,从而为评估和比较新方法提供了有用的测试套件。访问生成函数有助于理解为什么一种方法与其他算法相比表现不佳,从而提供改进的想法。包含大量文档和分析的资源是开源的,可在GitHub上获得。