Figure: Control human motion directly from latent space of Transformer-VAE.
Controlling 3D Human Action with Transformer Variational Autoencoder in Latent Space
In this repository, we present method for controlling human motion from previously developed Transformer-VAE model (ACTOR). With this method you can control human motion through discovered direction using low-rank factorization. We also proposed scheduling scheme, and data augmetation for learning latent space with more active units. We show some interesting results below.
The environment for this repository is the same as the ACTOR. Please refer to README in the ACTOR](https://github.com/Mathux/ACTOR) repository for the environment settings.
We referenced the repos below for the code.
This code is distributed under an MIT LICENSE.
Note that our code depends on other libraries, including SMPL, SMPL-X, PyTorch3D, and uses datasets which each have their own respective licenses that must also be followed.