3DGSI-Assessor: A Large-Scale Dataset and An LMM-based Method for 3D Gaussian Splatting Image Quality Assessment

2026-08-04

AI Summary

Existing image quality assessment (IQA) methods struggle to effectively characterize the unique distortions introduced by compressed 3D Gaussian splatting in both geometric and color dimensions, and typically yield only a single overall score without multidimensional diagnostic capability. To address this limitation, this work presents the first large-scale multidimensional subjective quality dataset for compressed 3D Gaussian splatting, named 3DGS-IEval-15K+, and introduces a unified evaluation framework based on large multimodal models (LMMs). By integrating global semantics with local dimensional features, the proposed method simultaneously predicts overall, geometric, and color quality scores in a single forward pass. Experiments demonstrate that the approach achieves state-of-the-art performance on the newly curated dataset and exhibits strong generalization in novel view synthesis tasks. The code and dataset are publicly released.

Abstract

3D Gaussian Splatting (3DGS) has become a dominant representation for real-time novel view synthesis (NVS), yet its storage footprint makes compression indispensable for practical deployment. 3DGS training and compression introduce representation-specific distortions such as floating artifacts and surface scattering, which conventional image quality assessment (IQA) metrics fail to capture. Moreover, the independent compression of geometric and color attributes may lead to decoupled dimension-specific distortions that must be diagnosed separately, yet existing metrics report only a single overall score. To address these gaps, we present 3DGS-IEval-15K+, a large-scale, multi-dimensional IQA dataset for compressed 3DGS, comprising 15,200 images from 10 diverse scenes, produced by 6 representative 3DGS algorithms at systematically designed compression levels and rendered from 20 strategically selected viewpoints spanning both training views and challenging novel views, annotated with 45,600 mean opinion scores (MOSs) across overall, geometry, and color quality. Based on 3DGS-IEval-15K+, we propose 3DGSI-Assessor, an all-in-one 3DGS IQA framework that integrates global semantic and dimension-specific local features within a large multimodal model (LMM), predicting all three dimensions in a single forward pass. 3DGSI-Assessor achieves state-of-the-art performance on 3DGS-IEval-15K+, and exhibits competitive generalization on other NVS benchmarks. Dataset and code will be released at https://github.com/YukeXing/3DGSI-Assessor.

Problem

• 3D Gaussian Splatting
• Image Quality Assessment
• Compression Artifacts
• Dimension-specific Distortions
• Novel View Synthesis

Innovation

• 3D Gaussian Splatting
• Image Quality Assessment
• Large Multimodal Model
• Multi-dimensional Evaluation
• Novel View Synthesis