3,950 remote sensing images featuring 11,033 annotated ship instances within port and offshore scenes. These medium-resolution satellite captures specifically target 'tiny' vessels that present unique identification challenges due to their small pixel footprint.
Use Cases
- Train object detection models to identify small maritime vessels using bounding box annotations
- Evaluate model recall for the 'tiny ship' class using the provided bounding box annotations
- Analyze the impact of 2.2-meter spatial resolution on the detection of small-scale maritime targets
Strengths
- Contains 3,950 images with a fixed resolution of 800x600 pixels
- Includes 11,033 ship instances annotated with bounding box coordinates
- Features medium-resolution remote sensing data with a spatial resolution of 2.2 meters