2023 Stanford FLAME AI Workshop

Click here to access the landing page of the 2023 Stanford FLAME AI Workshop.

This workshop (with 700 virtual participants) aimed to foster a dynamic forum for exchanging ideas, data, methods, and models related to ML techniques for fluid dynamics, turbulence, and combustion - fields crucial to the development of energy, propulsion, climate, and safety systems via a:

  1. A 10-day-long ML challenge (involving 1-2 person(s) per team) will be held to tackle generative modeling in fluid dynamics/turbulence with open-source data. Prize: GPU credits, and leading teams will be invited towards a joint publication.
  2. Daily talks on cutting-edge trends will be given by AI/ML experts within Stanford, academic guests, and industry partners from the Greater Silicon Valley ecosystem. Topics include:

2023 FLAME AI Challenge

We used the BLASTNet Momentum128 3D SR Dataset for 2023 FLAME AI Challenge. Click here to access the landing page of this challenge/competition.

Participants in this Kaggle challenge will be provided with a dataset containing low-resolution 2D flowfield images. Their task is to develop state-of-the-art super-resolution models that can reconstruct high-resolution versions of these flowfields. The challenge lies in preserving the fine details and structures in the flowfields while increasing the resolution, which is critical for maintaining the accuracy of simulations and analyses.

challenge