DIAL-GS: Dynamic Instance Aware Reconstruction for Label-free Street Scenes with 4D Gaussian Splatting

TL;DR

DIAL-GS introduces a self-supervised, instance-aware 4D Gaussian Splatting approach for detailed, dynamic urban scene reconstruction, improving accuracy and editing capabilities without relying on manual annotations.

Contribution

The paper presents a novel dynamic instance-aware reconstruction method using 4D Gaussian Splatting that accurately identifies and models dynamic objects in street scenes without supervision.

Findings

Outperforms existing self-supervised methods in reconstruction quality

Enables fine-grained instance-level editing of urban scenes

Enhances consistency and integrity through reciprocal dynamic perception

Abstract

Urban scene reconstruction is critical for autonomous driving, enabling structured 3D representations for data synthesis and closed-loop testing. Supervised approaches rely on costly human annotations and lack scalability, while current self-supervised methods often confuse static and dynamic elements and fail to distinguish individual dynamic objects, limiting fine-grained editing. We propose DIAL-GS, a novel dynamic instance-aware reconstruction method for label-free street scenes with 4D Gaussian Splatting. We first accurately identify dynamic instances by exploiting appearance-position inconsistency between warped rendering and actual observation. Guided by instance-level dynamic perception, we employ instance-aware 4D Gaussians as the unified volumetric representation, realizing dynamic-adaptive and instance-aware reconstruction. Furthermore, we introduce a reciprocal mechanism through which identity and dynamics reinforce each other, enhancing both integrity and consistency. Experiments on urban driving scenarios show that DIAL-GS surpasses existing self-supervised baselines in reconstruction quality and instance-level editing, offering a concise yet powerful solution for urban scene modeling.