Problem
Research questions and friction points this paper is trying to address.
Image-to-3D generation
Ultra-high resolution
Computational cost
Geometric detail
3D geometry
FILIGREE3D: Scaling Sparse Latent Flow Matching for Ultra-High-Resolution Image-to-3D Generation
🤖 AI Summary
This study addresses the inherent conflict between surging computational costs and detail degradation in ultra-high-resolution image-to-3D generation by proposing a sparse latent flow matching framework. Methodologically, it employs a sparse Diffusion Transformer (DiT) as the backbone, integrating local-global alternating attention with flow matching algorithms. Furthermore, the framework incorporates structure-aware sparse scaling, multi-scale feature injection, and visibility-aware volumetric regularization strategies. This work achieves single-image 3D generation at voxel resolutions up to 2048³ for the first time, synthesizing high-fidelity geometry within one minute on mainstream hardware. The proposed approach significantly outperforms existing baseline methods in both efficiency and reconstruction quality.
📝 Abstract
Scaling image-to-3D generation to ultra-high resolutions requires controlling rapidly growing computational costs without sacrificing fine geometric detail. We present \textbf{Filigree3D}, a sparse latent flow-matching framework that generates 3D geometry from a single image at voxel resolutions up to $2048^3$, with straightforward extensibility to $4096^3$. To make training tractable, we introduce Structure-Aware Sparse Scaling, which combines spatial bounding with alternating local-global attention to constrain token growth while preserving both fine-scale details and long-range structural context. To enhance detail reconstruction, we curate training samples based on their high-resolution geometric gains and inject multi-scale image features into a sparse 3D DiT, effectively coupling structural semantics with fine-grained visual cues. Furthermore, a visibility-aware voxel regularization strategy improves robustness against sparse perturbations and facilitates the completion of unobserved geometry. Under our default configuration, Filigree3D maintains peak GPU memory consumption within practical limits for contemporary hardware, enabling the generation of highly intricate 3D geometry in approximately one minute. Extensive experiments demonstrate that our method yields substantial improvements in overall geometric fidelity and fine-detail preservation compared to existing baselines, validating practical, detail-preserving 3D generation at unprecedented resolutions.
Innovation
Methods, ideas, or system contributions that make the work stand out.
Sparse Latent Flow Matching
Structure-Aware Sparse Scaling
Sparse 3D DiT
Image-to-3D Generation
Visibility-aware Voxel Regularization
🔎 Similar Papers
No similar papers found.
