Search papers, labs, and topics across Lattice.
This paper introduces Fake3DGS, a new benchmark for detecting manipulations in 3D Gaussian Splatting scenes, addressing the lack of research in 3D fake detection despite the increasing realism and ease of editing 3D content. The benchmark includes manipulated scenes with alterations to geometry, appearance, and spatial layout, rendered as realistic images. Experiments show that existing 2D detectors perform poorly on this task, but a novel 3D-aware detection method leveraging multi-view coherence and features from the Gaussian splatting representation significantly improves performance.
Current image forensics fall flat when faced with the subtle manipulations now possible in 3D Gaussian Splatting scenes, highlighting a critical gap in content authenticity assessment.
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.
