DAPlankton
收藏资源简介:
DAPlankton是一个用于多仪器浮游生物识别的精细领域适应基准数据集,由LUT大学计算机视觉与模式识别实验室创建。该数据集包含超过11万张通过不同仪器获取的浮游植物图像,分为DAPlanktonLAB和DAPlanktonSEA两个子集,分别代表实验室培养和自然环境中的浮游生物。数据集创建过程中,通过专家验证确保图像分类的准确性,旨在解决不同仪器间数据集的领域适应问题,为环境研究和海洋生物多样性分析提供支持。
DAPlankton is a fine-grained domain adaptation benchmark dataset for multi-instrument plankton recognition, developed by the Computer Vision and Pattern Recognition Laboratory of LUT University. This dataset contains over 110,000 phytoplankton images acquired by various instruments, and is divided into two subsets: DAPlanktonLAB and DAPlanktonSEA, which represent plankton in laboratory cultures and natural environments respectively. Expert validation was conducted during the dataset construction to ensure the accuracy of image classification, which aims to address the domain adaptation problem across datasets from different instruments and provide support for environmental research and marine biodiversity analysis.
- 1DAPlankton: Benchmark Dataset for Multi-instrument Plankton Recognition via Fine-grained Domain Adaptation计算机视觉与模式识别实验室, LUT大学, 芬兰 · 2024年



