CL-Splats: Continual Learning of Gaussian Splatting with Local Optimization

TL;DR

CL-Splats introduces an incremental method for updating 3D scene representations using Gaussian splatting, enabling efficient, high-quality reconstructions in dynamic environments through local optimization and change detection.

Contribution

It presents a novel approach combining change detection with local optimization for continual learning of 3D scenes using Gaussian splatting, improving efficiency and quality.

Findings

Achieves efficient scene updates with high reconstruction quality.

Outperforms existing methods in dynamic scene reconstruction.

Supports temporal scene analysis and state recovery.

Abstract

In dynamic 3D environments, accurately updating scene representations over time is crucial for applications in robotics, mixed reality, and embodied AI. As scenes evolve, efficient methods to incorporate changes are needed to maintain up-to-date, high-quality reconstructions without the computational overhead of re-optimizing the entire scene. This paper introduces CL-Splats, which incrementally updates Gaussian splatting-based 3D representations from sparse scene captures. CL-Splats integrates a robust change-detection module that segments updated and static components within the scene, enabling focused, local optimization that avoids unnecessary re-computation. Moreover, CL-Splats supports storing and recovering previous scene states, facilitating temporal segmentation and new scene-analysis applications. Our extensive experiments demonstrate that CL-Splats achieves efficient updates with improved reconstruction quality over the state-of-the-art. This establishes a robust foundation for future real-time adaptation in 3D scene reconstruction tasks.