MUGSQA: Novel Multi-Uncertainty-Based Gaussian Splatting Quality Assessment Method, Dataset, and Benchmarks

2025-11-10

AI Summary

Gaussian Splatting (GS) 3D reconstruction lacks reliable perceptual quality assessment methods, particularly under diverse input uncertainties. Method: This paper introduces the first subjective quality assessment framework for GS explicitly designed to handle multiple sources of uncertainty—view count, image resolution, observation distance, and initial point cloud accuracy. It innovatively proposes a multi-distance subjective evaluation protocol and constructs MUGSQA, the first benchmark dataset for GS quality assessment covering these uncertainties, accompanied by a dual-benchmark evaluation platform. Rigorous control of rendering and acquisition variables enables systematic robustness evaluation of mainstream GS methods under challenging conditions. Contribution/Results: The study exposes critical inconsistencies between conventional objective metrics (e.g., PSNR, LPIPS) and human perception. Experiments demonstrate that the proposed framework significantly enhances the credibility and practical utility of GS quality assessment.

Technology Category

• Computer Vision: 3D Computer Vision
• Machine Learning: Calibration & Uncertainty Quantification
• Knowledge Representation and Reasoning: Qualitative Reasoning

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Systems and Infrastructure for Web, Mobile and WoT: Web performance, measurement, and characterization

Abstract

Gaussian Splatting (GS) has recently emerged as a promising technique for 3D object reconstruction, delivering high-quality rendering results with significantly improved reconstruction speed. As variants continue to appear, assessing the perceptual quality of 3D objects reconstructed with different GS-based methods remains an open challenge. To address this issue, we first propose a unified multi-distance subjective quality assessment method that closely mimics human viewing behavior for objects reconstructed with GS-based methods in actual applications, thereby better collecting perceptual experiences. Based on it, we also construct a novel GS quality assessment dataset named MUGSQA, which is constructed considering multiple uncertainties of the input data. These uncertainties include the quantity and resolution of input views, the view distance, and the accuracy of the initial point cloud. Moreover, we construct two benchmarks: one to evaluate the robustness of various GS-based reconstruction methods under multiple uncertainties, and the other to evaluate the performance of existing quality assessment metrics. Our dataset and benchmark code will be released soon.

Problem

• Assessing perceptual quality of 3D objects reconstructed with Gaussian Splatting methods
• Evaluating GS reconstruction robustness under multiple input data uncertainties
• Measuring performance of existing quality assessment metrics for GS outputs

Innovation

• Multi-distance subjective quality assessment method
• Dataset considering multiple input uncertainties
• Benchmarks for reconstruction and quality metrics