The amount of available satellite imagery data has been substantially growing since the start of the Sentinel missions. Nevertheless, applications of AI to EO data are still scarce. The main goal of the AiTLAS project is to facilitate the uptake of EO data by AI experts and vice versa – the uptake of (advanced) AI methods by EO experts. This is achieved through the development of a comprehensive toolbox with resources such as: benchmarking tools, ready-to-exploit models, tools for learning models de novo, and semantically annotated datasets prepared in a format that is easy to use by AI methods.
The AiTLAS (Artificial Intelligence Toolbox for Earth Observation)t oolbox potential and usefulness is showcased by the execution of three pilots: a development of an EO Data benchmarking repository; a Maya archeological sites challenge; and crop type prediction for Slovenia, the Netherlands and Denmark.
AiTLAS targets two user groups: EO practitioners and AI practitioners. Considering the existing gap between the available EO imagery data and the developments in AI, AiTLAS is tailored for easy use by both user groups as follows. For EO practitioners, it provides easy-to-use interfaces to a variety of deep learning methods (through JSON file configurations and/or Jupyter notebooks). For AI practitioners, it provides easy-to-exploit EO data already implemented in AI-ready format.
The AiTLAS toolbox includes state-of-the-art machine learning methods for exploratory and predictive analysis of satellite imagery as well as a repository of AI-ready Earth Observation (EO) datasets. It can be easily applied for a variety of Earth Observation tasks, such as land use and land cover classification, crop type prediction, localization of specific objects (semantic segmentation), etc.
AiTLAS has several distinguishing properties: Maya archeological sites location;and crop type prediction for Slovenia, the Netherlands and Denmark.
- It is modular and flexible – allowing for easy configuration, implementation and extension of new data and models,
- It is general and applicable to a variety of tasks and workflows,
- It allows for fast and easy implementation of prototype solutions as well as implementation of complex analysis workflows.
- It is user-friendly.
In sum, AiTLAS, aides the AI community to engage in EO related tasks, by providing access to structured EO data, but more importantly, it facilitates and accelerates the uptake of (advanced) machine learning (AI) methods by the EO experts, thus bringing these two communities closer together.