DeepPhysX

Interfacing AI with simulation

The DeepPhysX project provides Python packages allowing users to easily interface their numerical simulations with learning algorithms.

DeepPhysX is mainly designed for SOFA and PyTorch frameworks, but other simulation and AI frameworks can also be used.

The project is closely linked to the SSD external Python library.

Let’s get started

To better understand the architecture of DeepPhysX, reading the above sections is highly recommended:

  • About → A quick summary of DeepPhysX features.

  • Overview → The project architecture and packages.

Then, to start using DeepPhysX, please refer to the Install section to configure your own DeepPhysX installation. Some examples are provided in each Example repository to learn how to use the package, with a 3 leveled complexity:

  • Tutorial: A walkthrough of DeepPhysX to learn the basics.

  • Features: A full session reviewing all features of DeepPhysX via a simple example.

  • Demos: Demonstration sessions with high-level applications.

Finally, to go further in understanding some components of DeepPhysX, refer to the associated section.