Cell-APP datasets

Datasets mentioned in the paper: "Cell-APP: A generalizable method for cell annotation and cell-segmentation model training." All folders in this repository follow the below outlined structure, with the exception of "general." -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- ZIP file structure:for_zenodo_d - dataset_name - annotations (folder containing annotated images; for visualization) - train - annotations (folder containing all individual masks) - images (folder containing microscopy images) - dataset_name.json (COCO dataset file) - test (folder with same structure as train) -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- In "general," you will find eight folders titled "dataset_N" where N varies from one to eight. These are the eight multi-cell line datasets of increasing size. In each folder titled "dataset_N," you will find a folder titled "data" containing microscopy images and a file titled "labels.json." labels.json is the COCO dataset file.论文《Cell-APP：一种可推广的细胞注释与细胞分割模型训练通用方法》中提及的数据集。 本仓库内除「general」文件夹外，其余所有文件夹均遵循下述目录结构。 ZIP文件结构（for_zenodo_d）： - dataset_name - annotations：用于存储供可视化使用的带注释图像 - train - annotations：存储所有独立分割掩码的文件夹 - images：存储显微图像的文件夹 - dataset_name.json：COCO格式数据集文件 - test：目录结构与train文件夹一致 在「general」文件夹中，包含8个命名为「dataset_N」的子文件夹，其中N的取值范围为1至8，对应8个尺寸逐步递增的多细胞系数据集。在每个「dataset_N」文件夹内，均设有一个名为「data」的子文件夹，用于存放显微图像，同时还包含一个名为「labels.json」的文件，该文件为COCO格式数据集文件。