SuperYOLO is a deep learning object detection system designed for remote sensing imagery. It combines YOLO-based object detection with a super-resolution branch to improve detection accuracy by processing high-resolution (1024x1024) images downsampled to 512x512. The model performs multimodal fusion using RGB and infrared channels and transforms labels for horizontal bounding box detection. It was published in IEEE Transactions on Geoscience and Remote Sensing.