30 April 2021
The new FAIR1M (Fine-grAined object recognItion in high-Resolution remote sensing imagery) database of 15,000 high-definition images with 1 million labelled ‘scenes’ and objects can aid AI’s accuracy, such as enabling it to identify not only a plane but its model type. All objects in the FAIR1M dataset are annotated with respect to 5 categories and 37 sub-categories by oriented bounding boxes. Compared with existing detection datasets dedicated to object detection, the FAIR1M dataset is much larger, it provides more rich fine-grained category information, it contains geographic information and it provides better image quality owing to a careful data cleaning procedure. The database was built with funding from the China National Science Foundation and access to the brand-new Gaofen observation satellites (80% of the data) and complemented with Google Earth (20 %). In May, AI researchers from many countries will compete in Beijing for a trophy awarded for satellite image recognition technology, using the FAIR1M database.
science paper: FAIR1M: A Benchmark Dataset for Fine-grained Object Recognition in High-Resolution Remote Sensing Imagery