AquaScan-1K Side Scan Sonar Dataset

AquaScan-1K Dataset AquaScan-1K is a specially curated side-scan sonar (SSS) dataset designed for underwater human detection in search-and-rescue scenarios. In total 1,000-high-resolution SSS images were acquired at various depths and across different types of lakebed using multiple scanning angles and operating frequencies ranging from 455 to 1075 kHz. Human surrogates were placed in controlled conditions during data acquisition. All images are annotated with axis-aligned bounding boxes for a single class ("human"), based on expert annotations provided by Red Cross personnel. Due to this setup, many targets appear as small, low-contrast objects. This makes AquaScan-1K a highly realistic and challenging reference dataset for small-object detection in sonar imagery. # Data structure The dataset is organized as follows: AquaScan-1K/├── images/├── labels/└── classes.txt images/Contains all side-scan sonar (SSS) images used in the dataset.Files are stored in standard image formats (e.g., .png or .jpg).Each image has a corresponding annotation file in the labels/ directory with the same file name (but .txt extension). labels/Contains the bounding-box annotations in YOLO format.For each image images/XYZ.png, there is a file labels/XYZ.txt.Each line in a label file follows the standard YOLO format: <class_id> <x_center> <y_center> <width> <height> where all coordinates are normalized to the range [0,1][0, 1][0,1] with respect to the image width and height. classes.txtDefines the class index mapping for the YOLO annotations.