GSCompleter: A Distillation-Free Plugin for Metric-Aware 3D Gaussian Splatting Completion in Seconds

2026-04-21

AI Summary

This work addresses the challenge of geometric holes and artifacts in 3D Gaussian Splatting (3DGS) under sparse-view settings, where existing completion methods rely on unstable iterative distillation pipelines. The authors propose a plug-and-play, distillation-free framework that enables efficient, metric-aware 3DGS reconstruction through a novel “generation–registration” paradigm. Key innovations include the first introduction of a Stereo-Anchor mechanism to generate metrically consistent 3D primitives, and a Ray-Constrained Registration strategy for global geometric fusion. By integrating 2D reference image synthesis, stereo-anchor-guided 3D lifting, and ray-constrained registration, the method substantially outperforms current approaches, achieving state-of-the-art completion quality and efficiency while consistently improving multiple baselines across three benchmarks.

Technology Category

• Computer Vision: 3D Computer Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Security and Privacy: Large-scale security measurements
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

While 3D Gaussian Splatting (3DGS) has revolutionized real-time rendering, its performance degrades significantly under sparse-view extrapolation, manifesting as severe geometric voids and artifacts. Existing solutions primarily rely on an iterative "Repair-then-Distill" paradigm, which is inherently unstable and prone to overfitting. In this work, we propose GSCompleter, a distillation-free plugin that shifts scene completion to a stable "Generate-then-Register" workflow. Our approach first synthesizes plausible 2D reference images and explicitly lifts them into metric-scale 3D primitives via a robust Stereo-Anchor mechanism. These primitives are then seamlessly integrated into the global context through a novel Ray-Constrained Registration strategy. This shift to a rapid registration paradigm delivers superior 3DGS completion performance across three distinct benchmarks, enhancing the quality and efficiency of various baselines and achieving new SOTA results.

Problem

• 3D Gaussian Splatting
• sparse-view extrapolation
• geometric voids
• artifacts
• scene completion

Innovation

• 3D Gaussian Splatting
• distillation-free
• scene completion
• Stereo-Anchor
• Ray-Constrained Registration