公开首个用于水下垃圾探测的前视声呐数据集，助力智能水下机器人发展。
The Marine Debris Forward-Looking Sonar Datasets
- 构建三类场景下的前视声呐数据集，提升样本多样性。
- 支持物体分类、检测、分割等五项视觉任务，覆盖多种学习范式。
- 适合水下机器人、海洋环境监测与声呐算法研究者使用。
声呐感知是水下机器人核心技术，但受限于人工智能系统的训练需求，公共声呐数据集匮乏。本文发布「Marine Debris Forward-Looking Sonar datasets」，包含水箱、转台和淹没采石场三种实验场景，增强数据多样性，并支持物体分类、目标检测、语义分割、图像匹配及无监督学习等多项计算机视觉任务。提供完整数据说明、基础分析及部分任务的初步结果。该数据集已公开，可从 https://doi.org/10.5281/zenodo.15101686 获取，期望推动相关领域研究进展。
原文摘要 · Abstract (English)
Sonar sensing is fundamental for underwater robotics, but limited by capabilities of AI systems, which need large training datasets. Public data in sonar modalities is lacking. This paper presents the Marine Debris Forward-Looking Sonar datasets, with three different settings (watertank, turntable, flooded quarry) increasing dataset diversity and multiple computer vision tasks: object classification, object detection, semantic segmentation, patch matching, and unsupervised learning. We provide full dataset description, basic analysis and initial results for some tasks. We expect the research community will benefit from this dataset, which is publicly available at https://doi.org/10.5281/zenodo.15101686
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。