This dataset is designed for obstacle detection tasks in visually impaired navigation scenarios.
The dataset is annotated in YOLO format and can be used for real-time object detection research on mobile or embedded devices. It is mainly intended for lightweight YOLO-based models such as YOLOv8.