D-MiSo: Editing Dynamic 3D Scenes using Multi-Gaussians Soup

TL;DR

D-MiSo introduces a novel method for editing dynamic 3D scenes by modeling Gaussian splats as a mesh-inspired structure, enabling flexible and partial scene modifications over time.

Contribution

The paper presents D-MiSo, a new approach that links Gaussian splats into a mesh-like structure for improved editing of dynamic scenes, overcoming previous limitations.

Findings

Allows separate trajectory construction for scene objects

Enables editing of dynamic scenes over time

Maintains partial scene dynamics during editing

Abstract

Over the past years, we have observed an abundance of approaches for modeling dynamic 3D scenes using Gaussian Splatting (GS). Such solutions use GS to represent the scene's structure and the neural network to model dynamics. Such approaches allow fast rendering and extracting each element of such a dynamic scene. However, modifying such objects over time is challenging. SC-GS (Sparse Controlled Gaussian Splatting) enhanced with Deformed Control Points partially solves this issue. However, this approach necessitates selecting elements that need to be kept fixed, as well as centroids that should be adjusted throughout editing. Moreover, this task poses additional difficulties regarding the re-productivity of such editing. To address this, we propose Dynamic Multi-Gaussian Soup (D-MiSo), which allows us to model the mesh-inspired representation of dynamic GS. Additionally, we propose a strategy of linking parameterized Gaussian splats, forming a Triangle Soup with the estimated mesh. Consequently, we can separately construct new trajectories for the 3D objects composing the scene. Thus, we can make the scene's dynamic editable over time or while maintaining partial dynamics.