flâneur

3D Gaussian Splatting as Markov Chain Monte Carlo

ubc-vision.github.io · 277 words · saved by 1 readers

Novel view reconstructions for (right) our method and (left) conventional 3D Gaussian Splatting with random initializations. Our method, even with random initialization, faithfully reconstructs the scene (e.g.. buildings at the back and the ground texture) providing much higher quality renderings. While 3D Gaussian Splatting has recently become popular for neural rendering, current methods rely on carefully engineered cloning and splitting strategies for placing Gaussians, which can lead to poor-quality renderings, and reliance on a good initialization. In this work, we rethink the set of 3D Gaussians as a random sample drawn from an underlying probability distribution describing the physical representation of the scene---in other words, Markov Chain Monte Carlo (MCMC) samples. Under this view, we show that the 3D Gaussian updates can be converted as Stochastic Gradient Langevin Dynamics (SGLD) update by simply introducing noise. We then rewrite the densification and pruning strategies

NeurIPS 2024 SPOTLIGHT 3D Gaussian Splatting as Markov Chain Monte Carlo 1University of British Columbia 2Google Research 3Google DeepMind 4Simon Fraser University 5University of Toronto Novel view reconstructions for (right) our method and (left) conventional 3D Gaussian Splatting with random initializations. Our method, even with random initialization, faithfully reconstructs the scene (e.g.. buildings at the back and the ground texture) providing much higher quality renderings. Abstract While 3D Gaussian Splatting has recently become popular for neural rendering, current methods…

saved by

related reading