sbatchelder/NES-plankton-classifier-2022-dataset

NES Plankton Classifier 2022 Training Data 是一个用于浮游生物物种图像分类的数据集，图像由美国东北部大陆架的成像流式细胞仪（IFCB）拍摄。该数据集包含155类浮游生物和非浮游生物ROI（包括碎屑、气泡、纤维等），共有97,026张标记图像，分为训练集和验证集。图像是从IFCB样本文件中提取的灰度PNG格式。数据集还包括类别标签、人类可读的类别名称和IFCB ROI持久标识符等元数据。

The NES Plankton Classifier 2022 Training Data is an image classification dataset of plankton species imaged by Imaging FlowCytobot (IFCB) on the Northeast US Shelf. It contains 155 classes of plankton and non-plankton ROIs (including detritus, bubbles, fibers, etc.), with a total of 97,026 labeled images across train and validation splits. The images are grayscale PNGs extracted from IFCB sample files. The dataset also includes metadata such as class labels, human-readable class names, and IFCB ROI persistent identifiers.