Back to Schedule
The ocean hides stories—some sailing proudly, others resting as silent wrecks beneath the waves. Detecting ships and maritime structures is essential for security, logistics, and even historical research, yet traditional methods often struggle with speed and scale. Furthermore, looking for ships using satellite images could be a great way to enhance your computer vision skills: this talk dives into how YOLO, a state-of-the-art object detection model that can revolutionize object monitoring, in this case by identifying ships and wrecks in real time from satellite and aerial imagery. We’ll explore the full pipeline: dataset preparation, annotation strategies for complex maritime scenes, and training YOLOv11 to handle challenges like varying ship sizes, cluttered backgrounds, and harsh weather conditions.