Gaussian Splatting: State-of-The-Arts and Future Trends

Birendra Kathariya, Dae Yeol Lee, Tsung-Wei Huang, Tong Shao, Peng Yin, Guan-Ming Su

Abstract: Novel view synthesis (NVS) from a sparse set of multi-view images has long been a challenging task in computer vision community. However, with the advent of Neural Radiance Field (NeRF) and 3D Gaussian Splatting (3DGS) technologies, a remarkable progress has been made in NVS and 3D scene representation. Especially, recent progress in 3DGS has unlocked new possibilities across various domains such as Augmented-Reality (AR)/Virtual-Reality (VR), sports, entertainment, computer graphics, and 3D modeling, sparking extensive research into its practical applications. In this article, we highlight the latest advances in 3DGS works focusing on representation, optimization, and compression. The goal of this paper is to explore state-of-the-art (SoTA) 3DGS works, offering insights into emerging trends and future directions in NVS and 3D scene representation.

External IDs:dblp:conf/iccnc/KathariyaLHS0S25