Table 2
Detail of layers of the encoder in DAEU network [23]. B is the number of channels of the input (spectrum). R is the number of units of the latent, hidden layer, i.e., the number of components to unmix. The utility layer performs an operation not specific to a neural network; in particular, the utility layer does not change the number of units.
| Layer# | Layer type | Activation | Units# |
|---|---|---|---|
| 1 | Input | – | B |
| 2 | Dense | LReLU | 9R |
| 3 | Dense | LReLU | 6R |
| 4 | Dense | LReLU | 3R |
| 5 | Dense | LReLU | R |
| 6 | Batch Normalization | Utility | R |
| 7 | Dynamical Soft Thresholding | LReLU | R |
| 8 | ASC enforcing | Utility | R |
| 9 | Gaussian Dropout | Utility | R |
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.