Python Deep Learning - Second Edition

1D, 2D, and 3D convolutions

Until now, we've used 2D convolutions, where the input and output neurons were arranged in a two-dimensional grid. This works very well for images. But we can also have 1D and 3D convolutions, where the neurons are arrange in one-dimensional or three-dimensional space respectively. In all cases, the filter has the same number of dimensions as the input and the weights are shared across the input. For example, we would use 1D convolution with time-series data, because the values are arranged across a single time axis. In the following diagram, we can see an example of 1D convolution:

1D convolution

The weights with ...