The project will use the ELSDSR database for training and testing, but we also might generate additional data with noise samples from the PRISM-set to increase accuracy.
For more information about the building and installing read the server and application readmes.