How to train a 1-d convolution VAE autoencoder on time series data ?

I am trying to implement

Variational Autoencoder (VAE) to Generate Images example

( https://www.mathworks.com/help/deeplearning/ug/train-a-variational-autoencoder-vae-to-generate-images.html) for time series data. I am having problem with understanding the dimensions of the sequence input which is in the form 'CBT' to add a fully connected layer to get mean and variances for desired dimensions.

[nick]

Hi Divya Kumawat,

I wanted to know whether you are refering 'CBT' form for your custom Encoder Decoder or the one specified in the documentation. The documentation example has 'CB' (Channel Batch) format and not CBT format.

I have attached the following link : https://www.mathworks.com/help/deeplearning/ref/dlarray.html for your reference to understand more about data formats used in DeepLearning Toolbox.