Con2DTranspose is an upsampling method used to increase the size of an image.
When we perform convolution, the size of the image decreases,
but in some scenarios, we want our image size to be the same as the input image size. Hence we use this convolution.
Here you will find Keras implementation on Conv2DTranspose
There are two "inputs" into Wavenet:
the previously generated samples of the waveform, which are usually encoded into multiple channels, like into 256 channels using 8-bit mu-law encoding
local conditioning, which can be things like linguistic features such as phoneme classes (used in the original wavenet paper), or frequencies like mel ...