3D Gaussian Splatting + Material Point Method (MPM)

Final project for CSCI 2240 Advanced Computer Graphics at Brown.

• We reimplemented PhysGaussian using Taichi.

• We also explored a simple extension combining PhysGaussian and System Idenfication, estimating the physical (mechanical) parameters (Young's Modulus) for 3D Gaussians.

Notes:

• Our implementation is built upon [1].

• We might clean the code a bit more and add more details for setup, usage, etc. in the future.

[1] Kerbl, B., Kopanas, G., Leimkühler, T., & Drettakis, G. (2023). 3d gaussian splatting for real-time radiance field rendering. ACM Transactions on Graphics, 42(4), 1-14.

[2] Xie, T., Zong, Z., Qiu, Y., Li, X., Feng, Y., Yang, Y., & Jiang, C. (2023). Physgaussian: Physics-integrated 3d gaussians for generative dynamics. CoRR abs/2311.12198 (2023).

[3] SIGGRAPH 2016 MPM Course

[4] MPM Implementation in Nvidia WARP

[5] Hu, Yuanming, et al. "Difftaichi: Differentiable programming for physical simulation." arXiv preprint arXiv:1910.00935 (2019).

[6] Li, Xuan, et al. "Pac-nerf: Physics augmented continuum neural radiance fields for geometry-agnostic system identification." arXiv preprint arXiv:2303.05512 (2023).

[7] Zhong, Licheng, et al. "Reconstruction and Simulation of Elastic Objects with Spring-Mass 3D Gaussians." arXiv preprint arXiv:2403.09434 (2024).