GoDe: Gaussians on Demand for Progressive Level of Detail and Scalable Compression

2025-01-23

AI Summary

To address the high storage overhead, excessive GPU memory consumption, lack of cross-device dynamic adaptability, and quality degradation from lossy compression in 3D Gaussian Splatting (3DGS), this paper proposes a hierarchical Gaussian organization architecture and a model-agnostic, progressive Level-of-Detail (LoD) mechanism. The LoD mechanism enables zero-shot, single-model adaptation—supporting multi-granularity compression and real-time detail control without retraining. Crucially, it achieves the first continuous, differentiable, and lossy compression integration for 3DGS. Evaluated on standard benchmarks, our method attains up to 10× compression ratio while maintaining reconstruction fidelity comparable to the original model. Both GPU memory footprint and bandwidth requirements become dynamically adjustable, significantly enhancing deployment flexibility and resource adaptability on edge devices.

Technology Category

• Computer Vision: 3D Computer Vision
• Machine Learning: Learning on the Edge & Model Compression
• Data Mining & Knowledge Management: Data Compression

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

3D Gaussian Splatting enhances real-time performance in novel view synthesis by representing scenes with mixtures of Gaussians and utilizing differentiable rasterization. However, it typically requires large storage capacity and high VRAM, demanding the design of effective pruning and compression techniques. Existing methods, while effective in some scenarios, struggle with scalability and fail to adapt models based on critical factors such as computing capabilities or bandwidth, requiring to re-train the model under different configurations. In this work, we propose a novel, model-agnostic technique that organizes Gaussians into several hierarchical layers, enabling progressive Level of Detail (LoD) strategy. This method, combined with recent approach of compression of 3DGS, allows a single model to instantly scale across several compression ratios, with minimal to none impact to quality compared to a single non-scalable model and without requiring re-training. We validate our approach on typical datasets and benchmarks, showcasing low distortion and substantial gains in terms of scalability and adaptability.

Problem

• 3D Gaussian Splattering
• Storage and Memory Efficiency
• Adaptability and Compression Artefacts

Innovation

• Multi-layer Detail Segmentation
• Adaptive Compression
• Flexible and Efficient Solution