computational optics + AI hackathon 🇨🇭
Chromatix is a differentiable wave optics library built with jax, which combines JIT-compilation, (multi-)GPU support, and automatic differentiation with a convenient programming style inspired by deep learning libraries. This makes Chromatix a great fit for anything requiring wave optics simulation, such as computational optics or solving inverse problems. More about the library can be found at the documentation page.
The developers of Chromatix are hosting a hackathon at the Janelia Research Campus (near DC), and we're organizing a satellite edition that will be hosted at the EPFL AI Center.
At EPFL, we will gather up to 20 engineers/researchers/scientists to work on various projects. You can work alone and in teams, and the Chromatix team will be available for in-person and remote support.
You can expect food, learning, and good times.
For questions, you can reach out to Eric Bezzam.
Tentative Program
The EPFL AI Center has kindly provided their Lounge (ELE 117) as a working space.
Tues May 27 (6pm):
Participant introductions
Introduction to Jax and Chromatix
May 28-30: hack, hack, hack
Note that Thurs May 29 is a holiday (Ascension)
You're not required to work non-stop and to be on-site all the time 😉 you're free to work on other stuff in between!
Sat May 31: Presentations and demos!
Project Ideas
Implementing your setup in Chromatix.
Using LLMs for building setups from a paper, writing documentation, etc.