Free Access

Table 2

Layout of the DCMDN architecture for the experiments with the quasar and the mixed catalog.

# Type Size # Maps Activation

1 input 15 × 16 × 16 / /
2 convolutional 3 × 3 256 tanh
3 pooling 2 × 2 256 tanh
4 convolutional 2 × 2 512 tanh
5 pooling 2 × 2 512 tanh
6 convolutional 2 × 2 1024 ReLu
7 MDN 500 / tanh
8 MDN 100 / tanh
9 output 15 / Eq. (2)

Notes. Stacks of 15 input images of the size 16 × 16 square-pixels are used.

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.