0
$\begingroup$

I am working on a problem of seq2seq modelling using ConvLSTM2D layer in keras. Implementation of convLSTM in keras allows user to control over output sequence using 'return_sequence' option. When True, the size of output sequence would be same as input sequence and when False, the output sequence would be a single frame. My question is how can I control on size of output sequence when it is neither same as input sequence nor a single frame (next frame)? For example, forecasting next 6 images from past 20 images using convLSTM2D. Any help in this regard would be helpful.

$\endgroup$

1 Answer 1

1
$\begingroup$

One solution would be to predict the next one, and now, based on 21 images you predict the second one, and so on and so forth.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.