Simple and Effective VAE Training with Calibrated Decoders

This is the master repo for the paper on Simple and Effective VAE Training with Calibrated Decoders. Check out the project page for the details!

Below is the list of released implementations, including σ-VAE in PyTorch, TensorFlow, and an implementation of σ-VAE on top of a sequential Stochastic Video Generation (SVG) model.

• [PyTorch GitHub]
• [PyTorch Colab]
• [TensorFlow GitHub]
• [TensorFlow Colab]
• [SVG]

Feel free to let me know if you have any questions by opening an issue or at [email protected]