Supporting Dataset for Localization and Classification of Gravity Wave Events from VIIRS Day/Night Band Satellite Imagery Using Machine Learning Techniques

This repository provides the supporting dataset for our study, Localization and Classification of Gravity Wave Events from VIIRS Day/Night Band Satellite Imagery Using Machine Learning Techniques. The dataset used for model training is included. It contains image files and text files that provide the corresponding labels. Dataset S1, uploaded as obj_data.zip, contains the data used for training, validation, and evaluation. We also provide the trained YOLOv8 weights file. Dataset S1: uploaded as “obj_data.zip”The ZIP file should be extracted after download. The resulting folder contains the PNG images and text files used for training. Each text file contains the label information for the corresponding image file. The format is: class_id, x_center_norm, y_center_norm, width_norm, height_norm Here, x_center_norm and y_center_norm indicate the center coordinates of the label in normalized image coordinates. width_norm and height_norm indicate the width and height of the label, also in normalized units. This is the annotation format supported by YOLOv8. The files train.txt, val.txt, and test.txt list the image filenames used for training, validation, and evaluation, respectively. The file data.yaml is the configuration file used for training and includes the class names and their corresponding class_id values. Weight S1: uploaded as “y8l_dnbgw.pt”This file contains the YOLOv8 network weights trained in this study. The file format is supported by YOLOv8.