Xiaoyuan Wang, Yizhou Zhao, Botao Ye, Xiaojun Shan, Weijie Lyu, Lu Qi, Kelvin C. K. Chan, Yinxiao Li, Ming-Hsuan Yang
HoliGS is a new framework for efficiently reconstructing dynamic environments from long monocular RGB videos using deformable Gaussian splatting, offering superior reconstruction quality and reduced training times compared to existing methods.
Researchers have developed HoliGS, a new method for creating 3D models of dynamic environments from long video recordings. This method uses a technique called Gaussian splatting to efficiently represent both static and moving parts of a scene. By breaking down scenes into static backgrounds and moving objects, HoliGS can handle complex scenarios with multiple actors and large viewpoint changes. The results show that HoliGS not only improves the quality of the 3D reconstructions but also speeds up the training and rendering processes compared to previous techniques.
