Utilization of transfer learning and data augmentation (rotation, flipping). 4. Experimental Results
1,061 images featuring over 28 categories of ships. sc24016-VTv08731.rar
Ship detection in high-resolution satellite imagery is critical for maritime safety and surveillance. This paper evaluates state-of-the-art object detection models using the HRSC2016 dataset, focusing on the challenges of multi-scale targets and complex harbor backgrounds. 1. Introduction flipping). 4. Experimental Results 1
Benchmark deep learning architectures on the HRSC2016 dataset . 2. Dataset Overview sc24016-VTv08731.rar
Evaluated using Mean Average Precision (mAP) and F1-score.
This file likely refers to the (High-Resolution Ship Collections 2016) dataset, a widely used benchmark for ship detection in remote sensing imagery . Paper Draft: Ship Detection Using HRSC2016
The dataset is available on platforms like IEEE DataPort and Roboflow Universe . To help me customize this paper for you: