DetailGen3D: Generative 3D Geometry Enhancement via Data-Dependent Flow

TL;DR

DetailGen3D is a novel method that enhances the geometric detail of 3D shapes generated from sparse or single views by modeling local refinements through data-dependent flows, improving detail fidelity efficiently.

Contribution

It introduces a data-dependent flow approach for 3D shape refinement, enabling detailed enhancement without large-scale 3D generative models.

Findings

Achieves high-fidelity geometric detail synthesis.

Maintains efficiency in training and inference.

Effective across various 3D generation methods.

Abstract

Modern 3D generation methods can rapidly create shapes from sparse or single views, but their outputs often lack geometric detail due to computational constraints. We present DetailGen3D, a generative approach specifically designed to enhance these generated 3D shapes. Our key insight is to model the coarse-to-fine transformation directly through data-dependent flows in latent space, avoiding the computational overhead of large-scale 3D generative models. We introduce a token matching strategy that ensures accurate spatial correspondence during refinement, enabling local detail synthesis while preserving global structure. By carefully designing our training data to match the characteristics of synthesized coarse shapes, our method can effectively enhance shapes produced by various 3D generation and reconstruction approaches, from single-view to sparse multi-view inputs. Extensive experiments demonstrate that DetailGen3D achieves high-fidelity geometric detail synthesis while maintaining efficiency in training.