\textit{4DSurf}: High-Fidelity Dynamic Scene Surface Reconstruction

2026-03-30

AI Summary

This work addresses the challenge of temporal geometric inconsistency in existing Gaussian splatting methods when handling large deformations and multi-object motion in dynamic scenes. The authors propose 4DSurf, a unified framework for dynamic surface reconstruction that requires no prior knowledge of object count or category. Its key innovations include regularizing motion trajectories via a signed distance function flow guided by Gaussian deformations and employing an overlapping segmentation strategy to propagate geometric information across frames, thereby preserving temporal consistency even under substantial deformations. Experiments demonstrate that 4DSurf outperforms state-of-the-art methods by 49% and 19% in Chamfer distance on the Hi4D and CMU Panoptic datasets, respectively, while significantly enhancing reconstruction stability and fidelity under sparse-view settings.

Abstract

This paper addresses the problem of dynamic scene surface reconstruction using Gaussian Splatting (GS), aiming to recover temporally consistent geometry. While existing GS-based dynamic surface reconstruction methods can yield superior reconstruction, they are typically limited to either a single object or objects with only small deformations, struggling to maintain temporally consistent surface reconstruction of large deformations over time. We propose ``\textit{4DSurf}'', a novel and unified framework for generic dynamic surface reconstruction that does not require specifying the number or types of objects in the scene, can handle large surface deformations and temporal inconsistency in reconstruction. The key innovation of our framework is the introduction of Gaussian deformations induced Signed Distance Function Flow Regularization that constrains the motion of Gaussians to align with the evolving surface. To handle large deformations, we introduce an Overlapping Segment Partitioning strategy that divides the sequence into overlapping segments with small deformations and incrementally passes geometric information across segments through the shared overlapping timestep. Experiments on two challenging dynamic scene datasets, Hi4D and CMU Panoptic, demonstrate that our method outperforms state-of-the-art surface reconstruction methods by 49\% and 19\% in Chamfer distance, respectively, and achieves superior temporal consistency under sparse-view settings.

Problem

• dynamic scene reconstruction
• surface reconstruction
• temporal consistency
• large deformations
• Gaussian Splatting

Innovation

• Gaussian Splatting
• Signed Distance Function Flow Regularization
• Overlapping Segment Partitioning
• Dynamic Surface Reconstruction
• Temporal Consistency