Skip to content

cdglissov/recurrent-flows-msc

Repository files navigation

recurrent-flows-msc

This repository contains the following implementations:

ConvLSTM
RealNVP
Glow
VRNN
SRNN with smoothing, residual mean and overshooting
Hierarchical SRNN (Not included yet)
Conditional Glow
SVG
Recurrent Flow Networks based on SRNN with smoothing, residual mean, gradient checking and overshooting and a Glow decoder - (Video generation)

Some examples can be found in the notebook folder.

Data

KTH
BAIR
SM-MNIST
Modified Celeba data for conditional experiments
Generated moving shapes data
Synthesized sinusoidal data
Two-moon data with rotation

About

Combining normalizing flows and deep latent variable models to generate video.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 3

  •  
  •  
  •