S3 Space ML

The Structure Detection on Large Scape Maps with Machine Learning service delivers a user-friendly cloudbased solution for innovative structure detection (e.g., compact/extended sources, filaments), extended the popular CAESAR/ CuTEx tools with machine learning frameworks.

The currently experienced dramatic increases of Space data volumes make automatic structure detection a necessity. This need will become more and more important due to the continuous increase of data volumes that will need to be analysed. Recently project partners were involved in projects for automatic structure recognition such as FP7 ViaLactea project and SKA precursor activities. CAESAR (Riggi et al. 2016) is a software tool for extraction and parameterization of both compact and extended sources present in astronomical maps. CuTEx (Molinari et al. 2011, Schisano et al. 2014) analyses images in the infrared bands and, in particular, it was designed to resolve problems concerning the study of star forming regions. Planetary exploration missions have the constant need for terrain characterization, largely based on orbital remote sensing data. S3 service will be focused on exploiting the existing TRL6 software to perform pattern and structure detection in astronomical surveys as well as in planetary surface composition, topography and morphometry. The service is expected to integrate cutting-edge machine learning algorithms, adopting pre-trained convolutional neural-networks for computer vision tasks (e.g., recognition, segmentation) adapted to the project-specific tasks by means of transfer learning approaches, to perform advanced classification for structures of sources in the sky or planetary surfaces to identify regions of interest.

Learn more about the service and all the technical details.

Discover more with a video presentation on our YouTube Channel.

 

EU Flag  NEANIAS is a Research and Innovation Action funded by European Union under Horizon 2020 research and innovation programme via grant agreement No.863448.