Temporal Difference Variational Auto-Encoder (TD-VAE) (Implementation in PyTorch)

TD-VAE in PyTorch

Mitigating the Filter Bubble while Maintaining Relevance: Targeted Diversification with VAE-based Recommender Systems

The code implements the framework proposed in Temporal Difference Variational Auto-Encoder (Gregor et al). The main purpose is to make it a simple block that is easy to use as a plugin of the network. As the sampling process in the multi-stochastic-layer version is extremely complicated, I try the trick of recursive function call to simplify it.

Simulator for different representation learning methods (e.g. Witness Autoencoder, Topological Autoencoder etc.)

Implementation of TD VAE from Gregor et al 2018

No description available

No description available

Synthesizing future brain states from longitudinal Magnetic Resonance Imaging

No description available

Using a temporal-difference variational autoencoder to predict abnormal cell division phenotypes

No description available

No description available

The implementation of VAE-TDW for PDFM

Implementation of the TD-VAE model (Gregor et. al, 2018)

NNDL Project on implementing TD-VAE with a Markovian latent prior for structured sequence modeling.

A Preliminary submission to the Lossfunk research internship assignment dealing with the creation of world-models purely from observing RL agents train using TD-VAE and keypoint tracking.