Search papers, labs, and topics across Lattice.
This paper introduces GeAR, a novel two-stage framework designed to reconstruct 3D representations from classical paintings by ensuring geometric plausibility and appearance fidelity. By converting paintings into geometry-grounded representations and restoring artwork-faithful appearances, GeAR addresses the challenges posed by the stylized and single-view nature of classical art. Extensive evaluations demonstrate that GeAR significantly outperforms existing methods in terms of both geometric accuracy and visual quality, as validated by user studies and a new benchmark dataset of over 10,000 artworks.
GeAR achieves unprecedented fidelity in reconstructing 3D scenes from classical paintings, outperforming traditional methods in both geometry and appearance.
Classical paintings preserve rich spatial, cultural, and historical content, making their reconstruction as explorable 3D scenes valuable for digital preservation, immersive exhibition, and cultural engagement. Yet, unlike photographs, they often depict scenes in a single-view, stylized manner, with weak perspective, lighting, and depth cues. Existing 3D reconstruction methods are largely built on natural-image priors, making it difficult to recover geometrically plausible and visually faithful 3D representations from such inputs. To address this challenge, we introduce Classical Painting-to-3D (CP3D), a new task that aims to recover a 3D representation from a single classical painting while jointly ensuring geometric plausibility, appearance fidelity to the source artwork, and plausible novel-view synthesis. We further propose GeAR, a training-free two-stage framework for Geometry Grounding and Appearance Restitution. GeAR first converts the input painting into a geometry-grounded representation with more coherent shading and illumination cues, improving the stability of 3D Gaussian reconstruction. It then restores artwork-faithful appearance across views under spatial constraints and multi-view consistency, recovering the painterly textures and details weakened during grounding. In addition, we construct HeriArch, a curated benchmark of 10,160 high-resolution classical artworks for systematic evaluation of CP3D. Extensive experiments and user studies show that GeAR consistently outperforms strong baselines in geometric plausibility, appearance fidelity, and human preference. Code and dataset will be released publicly.
