SplatCo: Structure-View Collaborative Gaussian Splatting for Detail-Preserving Rendering of Large-Scale Unbounded Scenes

2025-05-23

AI Summary

Addressing the challenge of high-fidelity rendering for large-scale, unbounded outdoor scenes—specifically, preserving global structural consistency without sacrificing local geometric and textural detail—this paper proposes a structure-view-cooperative Gaussian splatting framework. Methodologically, it introduces (1) a cross-structure collaboration module that jointly models geometry and appearance by integrating triplane representations with hierarchical contextual grids; and (2) a cross-view cooperative training strategy featuring visibility-aware adaptive densification/pruning and multi-view gradient synchronization. Evaluated on 13 large real-world scenes, the method consistently outperforms state-of-the-art approaches: PSNR improves by 1–2 dB, SSIM by 0.1–0.2, and it establishes, for the first time, a new benchmark for high-fidelity rendering in unbounded outdoor environments.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Planning, Routing, and Scheduling: Replanning and Plan Repair
• Humans and AI: Human-Aware Planning and Behavior Prediction

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Security and Privacy: Large-scale security measurements

Abstract

We present SplatCo, a structure-view collaborative Gaussian splatting framework for high-fidelity rendering of complex outdoor environments. SplatCo builds upon two novel components: (1) a cross-structure collaboration module that combines global tri-plane representations, which capture coarse scene layouts, with local context grid features that represent fine surface details. This fusion is achieved through a novel hierarchical compensation strategy, ensuring both global consistency and local detail preservation; and (2) a cross-view assisted training strategy that enhances multi-view consistency by synchronizing gradient updates across viewpoints, applying visibility-aware densification, and pruning overfitted or inaccurate Gaussians based on structural consistency. Through joint optimization of structural representation and multi-view coherence, SplatCo effectively reconstructs fine-grained geometric structures and complex textures in large-scale scenes. Comprehensive evaluations on 13 diverse large-scale scenes, including Mill19, MatrixCity, Tanks&Temples, WHU, and custom aerial captures, demonstrate that SplatCo consistently achieves higher reconstruction quality than state-of-the-art methods, with PSNR improvements of 1-2 dB and SSIM gains of 0.1 to 0.2. These results establish a new benchmark for high-fidelity rendering of large-scale unbounded scenes. Code and additional information are available at https://github.com/SCUT-BIP-Lab/SplatCo.

Problem

• High-fidelity rendering of large-scale unbounded scenes
• Preserving fine details in complex outdoor environments
• Enhancing multi-view consistency in scene reconstruction

Innovation

• Combines global tri-plane with local grid features
• Uses hierarchical compensation for detail preservation
• Synchronizes gradient updates for multi-view consistency