Radiance Field
Radiance fields are neural representations of 3D scenes that enable novel view synthesis and other advanced capabilities. Current research focuses on improving efficiency and realism, particularly through the use of Gaussian splatting, which offers faster rendering and better handling of view-dependent effects, as well as addressing challenges like handling dynamic scenes, inconsistent lighting conditions, and limited data. These advancements are significant for applications in robotics, virtual and augmented reality, and computer graphics, offering more realistic and efficient 3D scene modeling and manipulation.
Papers
SO-NeRF: Active View Planning for NeRF using Surrogate Objectives
Keifer Lee, Shubham Gupta, Sunglyoung Kim, Bhargav Makwana, Chao Chen, Chen Feng
Feature 3DGS: Supercharging 3D Gaussian Splatting to Enable Distilled Feature Fields
Shijie Zhou, Haoran Chang, Sicheng Jiang, Zhiwen Fan, Zehao Zhu, Dejia Xu, Pradyumna Chari, Suya You, Zhangyang Wang, Achuta Kadambi