Fake3DGS: A Benchmark for 3D Manipulation Detection in Neural Rendering

2026-04-30

AI Summary

Existing methods struggle to effectively detect 3D manipulations in neural renderings based on 3D Gaussian splatting, particularly concerning geometry, appearance, and spatial layout, posing significant authenticity and security risks. This work formally defines the task of 3D forgery detection and introduces Fake3DGS, the first benchmark dataset specifically designed for evaluating manipulations in 3D Gaussian splatting. Furthermore, we propose a 3D-aware detection framework that integrates multi-view consistency with intrinsic features of the Gaussian representation. Experimental results demonstrate that conventional 2D detectors exhibit limited performance on this task, whereas our approach substantially improves detection accuracy, thereby validating both the necessity of 3D-aware forgery detection and the effectiveness of the proposed dataset.

Abstract

Recent advances in 3D reconstruction and neural rendering,particularly 3D Gaussian Splatting, make it feasible and simple to edit 3D scenes and re-render them as highly realistic images. Therefore, security concerns arise regarding the authenticity of 3D content. Despite this threat, 3D fake detection remains largely unexplored in the literature, and most existing work is limited to 2D space. Therefore, in this paper, we formalize the concept of 3D fake detection and introduce Fake3DGS, a dataset of 3D Gaussian splatting scenes and corresponding rendered views, where fake images are produced by controlled manipulations of geometry, appearance, and spatial layout, while preserving high visual realism. Using this benchmark, we demonstrate that current state-of-the-art 2D detectors struggle to distinguish between original and 3D manipulated images. To bridge this gap, we introduce a 3D-aware detection method that leverages multi-view coherence and features derived from the Gaussian splatting representation. Experimental results demonstrate a substantial improvement in recognizing modified 3D content, underscoring the validity of the new dataset and the necessity for authenticity assessment techniques that extend beyond 2D evidence. Code and data are publicly released for future investigations.

Problem

• 3D manipulation detection
• neural rendering
• 3D Gaussian Splatting
• fake detection
• 3D authenticity

Innovation

• 3D fake detection
• 3D Gaussian Splatting
• multi-view coherence
• neural rendering
• manipulation benchmark