SINC: Spatial Composition of 3D Human Motions for Simultaneous Action Generation

ICCV 2023

Nikos Athanasiou*1, Mathis Petrovich*1,2,
Michael J. Black1, Gül Varol2

1MPI for Intelligent Systems, Tübingen, Germany
2LIGM, École des Ponts, Univ Gustave Eiffel, CNRS, France

Download SINC synthetic data, if you want to train your models with spatial composition from AMASS.
The data you are exploring in this demo are the data created using the code to compose motions from AMASS in our repo.**

Example Actions
Example Actions
Number of motions to be shown from the SINC-Synthetic data

**Our data in the official paper are using on the fly compositions, which means than are not computed and filtered offline. This is a minimally processed version of ~124k motions ranging between 3-7 seconds.

Made with ❤️ by Nikos Athanasiou