From NeRF to 3DGS: Exploring Advanced Methods for High-Quality Static and Dynamic Scene Reconstruction in Interactive 3D Worlds

In this training lab, we'll explore the latest techniques in 3D scene creation using neural radiance fields (NeRF) and 3D Gaussian splatting (3DGS). These methods use machine learning to create detailed 3D scenes from just a few images. Learn how to model dynamic scenes where objects can move and rotate while keeping their color, opacity, and size consistent. We’ll also cover how to combine elements from different scenes and add new objects to moving scenes, like attaching a hat to a character that changes position. Gain practical skills in using NeRF and 3DGS to create and manage both static and dynamic 3D scenes.
Prerequisite(s):

Python.