Segment Any 4D Gaussians

TL;DR

SA4D introduces a novel framework for segmentation in 4D Gaussian representations, enabling efficient, high-quality segmentation and manipulation of 4D scenes, advancing the understanding and editing of dynamic 4D environments.

Contribution

The paper presents SA4D, the first framework for 4D Gaussian segmentation, incorporating a temporal identity feature field and a refinement process for artifact removal.

Findings

Achieves precise 4D segmentation within seconds.

Enables removal, recoloring, and composition of 4D masks.

Demonstrates high-quality rendering of segmented 4D scenes.

Abstract

Modeling, understanding, and reconstructing the real world are crucial in XR/VR. Recently, 3D Gaussian Splatting (3D-GS) methods have shown remarkable success in modeling and understanding 3D scenes. Similarly, various 4D representations have demonstrated the ability to capture the dynamics of the 4D world. However, there is a dearth of research focusing on segmentation within 4D representations. In this paper, we propose Segment Any 4D Gaussians (SA4D), one of the first frameworks to segment anything in the 4D digital world based on 4D Gaussians. In SA4D, an efficient temporal identity feature field is introduced to handle Gaussian drifting, with the potential to learn precise identity features from noisy and sparse input. Additionally, a 4D segmentation refinement process is proposed to remove artifacts. Our SA4D achieves precise, high-quality segmentation within seconds in 4D Gaussians and shows the ability to remove, recolor, compose, and render high-quality anything masks. More demos are available at: https://jsxzs.github.io/sa4d/.