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

TL;DR

This paper introduces Fake3DGS, a benchmark dataset for detecting manipulated 3D scenes in neural rendering, revealing current detectors' limitations and proposing a 3D-aware detection method that improves authenticity assessment.

Contribution

The paper formalizes 3D fake detection, creates a new dataset with manipulated 3D scenes, and proposes a 3D-aware detection method to improve authenticity verification.

Findings

State-of-the-art 2D detectors struggle with 3D manipulated images.

The proposed 3D-aware method significantly improves detection accuracy.

The dataset enables future research in 3D content authenticity assessment.

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.