Dr.TVAM

ACM Transactions on Graphics (Proceedings of SIGGRAPH Asia), December 2024.
Baptiste Nicolet · Felix Wechsler · Jorge Madrid-Wolff · Christophe Moser · Wenzel Jakob

About this project

Dr.TVAM is a high-performance inverse rendering framework for tomographic volumetric additive manufacturing. It is based on the Mitsuba renderer, and uses physically-based differentiable rendering to optimize patterns for TVAM. In particular, it supports:

• Printing in scattering media
• Arbitrary vial shapes (round, square, ...)
• Arbitrary projector motions (orthogonal, tilted)
• An improved discretization scheme for the target shape to reduce computational load and to minimize discretization artifacts

For more details we refer to this publication.

⚠️ Sign up for online workshop in Spring 2025

• The date is not announced yet, but sign up for the announcement: here

Installation

Installing Dr.TVAM can be done via pip:

pip install drtvam

Basic Usage

We provide a convenience command-line tool drtvam to run simple optimizations. You can run it as:

drtvam path/to/config.json

Please refer to the documentation for details on the configuration file format. Dr.TVAM will run multi-threaded on your machine but will also use your CUDA GPU and your RT cores if supported by your hardware.

Advanced Usage

Dr.TVAM provides a set of useful abstractions to implement a wide variety of custom TVAM setups. We show examples in the documentation to get you started.

Documentation

The full documentation for this project, along with jupyter notebooks explaining the basics of implementing your own optimizations in our framework, can be found on readthedocs.

Teaser video

Can be found on YouTube:

Issues or support needed

In case you run into issues or you do need support, do not hesitate to open an issue such that we can help you using Dr. TVAM! As an academic user, this is completely free to use. Please reach out to us in case you need support!

License

This project is provided under a non-commercial license. Please refer to the LICENSE file for details.

Citation

When using this project in academic works, please cite the following paper:

@article{nicolet2024inverse,
    author = {Nicolet, Baptiste and Wechsler, Felix and Madrid-Wolff, Jorge and Moser, Christophe and Jakob, Wenzel},
    title = {Inverse Rendering for Tomographic Volumetric Additive Manufacturing},
    journal = {Transactions on Graphics (Proceedings of SIGGRAPH Asia)},
    volume = {43},
    number={6},
    year = {2024},
    month = dec,
    doi = {10.1145/3687924}
}