Datasets for: Integrating Automated Detection and Segmentation for Quantitative Analysis of Stomata and Pavement Cells using StomataQuant

This repository contains the supplementary datasets for the manuscript: "Integrating Automated Detection and Segmentation for Quantitative Analysis of Stomata and Pavement Cells using StomataQuant". Datasets Overview: - Supplementary Dataset S1 (Development Sets): Comprises training and validation datasets for three deep learning model: stomata detection (7,884 images), stomata and pore segmentation (500 images), and stomata and pavement cell segmentation (613 images). Each subset includes high-quality microscopy images and their corresponding YOLO-format annotations. - Supplementary Dataset S2 (Independent Test Sets): Features independent test datasets used to evaluate the cross-species generalizability of the StomataQuant models. It includes 300 images from 144 species for detection, and 45 images each for pore and pavement cell segmentation tasks, complete with ground-truth labels. - Supplementary Dataset S3 (Inference Time Assay Data): A dedicated dataset used to measure the computational efficiency of the automated analysis pipeline across preprocessing, model inference, and post-processing stages. Software Availability: The source code and graphical user interface (GUI) of StomataQuant are available at: https://github.com/Milo-L/StomataQuant