ml-for-3d-course/main/en/unit4/hands-on.md

Hands-on

Initially, we planned to walk through the Marching Cubes algorithm and apply it to the LGM Demo. However, recent advancements in mesh generation have made this approach less relevant.

While a deep dive into the methods behind MeshAnything would be much more pertinent, its newness and non-commercial license make it suboptimal for the time being.

Instead, here are some resources based on your goals:

• Splat to Mesh: If you followed along with with the LGM-based activities and want to produce the final mesh, this open source demo is based on the original LGM codebase. Note that this method is slow and resource-intensive.
• InstantMesh: This is fast and state-of-the-art approach uses FlexiCubes to produce the final mesh. It currently ranks toward the top of the 3D Arena leaderboard.
• meshgpt-pytorch: This open source reimplementation of MeshGPT provides a good starting point for open-source differentiable mesh generation. MeshAnything builds upon MeshGPT. Note: This implementation only provides architecture, not weights.

These resources should help you continue exploring mesh generation and its most recent advancements.