Fig. 3.

Representation of the first convolution layer with two kernels of convolution to simplify the schema. The multispectral image of a galaxy in the five SDSS filters is fed to the network. Each kernel is convolved with the multichannel images. The convolved images are summed, an additive bias is added and a non-linear function is applied, to produce one feature map.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.