January 2018
Intermediate to advanced
474 pages
10h 23m
English
These layers implement the different type of convolution, sampling, and cropping operations for convolutional neural networks:
| Layer Name | Description |
| Conv1D | This layer applies convolutions over a single spatial or temporal dimension to the inputs. |
| Conv2D | This layer applies two-dimensional convolutions to the inputs. |
| SeparableConv2D | This layer applies a depth-wise spatial convolution on each input channel, followed by a pointwise convolution that mixes together the resulting output channels. |
| Conv2DTranspose | This layer reverts the shape of convolutions to the shape of the inputs that produced those convolutions. |
| Conv3D | This layer applies three-dimensional convolutions to the inputs. |
| Cropping1D | This ... |
Read now
Unlock full access