Fig. 3.

CNN architecture. As explained in the text, it consists of five convolution layers composed of 64, 128, 128, 128, and 64 channels, including convolution, batch normalization, leaky ReLU activation (first three columns in light blue), maximum pooling, and Gaussian dropout (last two columns in dark blue). The last layer is followed by a fully connected layer (purple) and a classification layer returning a binary answer (1 for moon, 0 otherwise).
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.