Structure Detection with Machine Learning Techniques
The Space-ML services provide advanced solutions for pattern and structure detection in astronomical surveys as well as in planetary surface composition, topography and morphometry. The service includes tools such as CAESAR and integrates cutting-edge
Machine Learning algorithms to perform advanced classification for structures of sources in the sky or planetary surfaces to identify regions of interest.
CAESAR Service
The CAESAR Service performs automated source finding in astronomical maps. It allows to extract and parametrize both compact and extended sources from astronomical radio interferometric maps. The processing pipeline is a series of stages that can run
on multiple cores and processors. Compact sources are extracted with flood-fill and blob finder algorithms, processed and fitted using a 2D gaussian mixture model. Extended source search is based on a pre-filtering stage, allowing
image denoising, compact source removal and enhancement of diffuse emission, followed by a final segmentation. Different algorithms are available for image filtering and segmentation. The outputs delivered to the user include
source fitted and shape parameters, regions and contours.
The CAESAR Service is provided by INAF-Istituto Nazionale di Astrofisica, see its Terms of use and Privacy Policy.
Machine/Deep Learning Service
The Machine/Deep Learning Service is delivering innovative solutions to improve source identification, classification and characterisation in large-scale radio surveys.
The AstroML Service is provided by INAF-Istituto Nazionale di Astrofisica, see its Terms of use.
Latent Space Explorer
Latent Space Explorer (LSE) support analysis of image datasets via unsupervised machine learning methods.
The LSE Service is provided by UNIMIB-Università degli studi di Milano-Bicocca, see its Terms of use and Privacy Policy..
Read more on the
Space-ML Documentation