MOC-3D: Manifold-Order Consistency for Text-to-3D Generation

2026-05-03

AI Summary

Existing text-to-3D generation methods often suffer from macroscopic topological errors—such as the Janus effect—and microscopic geometric discontinuities, primarily due to view-dependent bias and gradient noise under high classifier-free guidance (CFG) scales. To address these issues, this work proposes MOC-3D, which introduces two key innovations within the ScaleDreamer framework: a viewpoint sequencing constraint grounded in CLIP semantic priors to enforce global structural consistency, and a feature continuity mechanism based on Riemannian metrics over symmetric positive-definite manifolds to ensure smooth local geometry. These components operate synergistically to enable high-quality 3D generation with multi-scale consistency, significantly mitigating topological artifacts while enhancing both structural and textural continuity across multiple viewpoints.

Abstract

With the burgeoning development of fields such as the Metaverse, Virtual Reality (VR), and Digital Twins, text-to-3D generation has emerged as a research hotspot in both academia and industry. Currently, optimization methods based on Score Distillation Sampling (SDS) utilizing 2D diffusion priors have become the mainstream technological paradigm in this field. However, due to the view bias of 2D priors and the mode-seeking ambiguity combined with gradient noise induced by high Classifier-Free Guidance (CFG), these methods still suffer from macro-topological inconsistency (e.g., the Janus problem) and micro-geometric discontinuity. To address these challenges, we propose MOC-3D, a text-to-3D generation method based on geometric manifold and semantic view-order consistency. Built upon the ScaleDreamer framework, our method incorporates a Semantic View-Order Constraint Module and a Manifold-based Feature Continuity Module. The former aims to rectify macro-topological inconsistency, while the latter focuses on eliminating micro-geometric discontinuity. Specifically, the Semantic View-Order Constraint Module leverages the prior knowledge of CLIP to impose a Monotonicity Rank Constraint on semantic score representations across different views, thereby providing effective guidance for the global topological structure of 3D objects. Meanwhile, the Manifold-based Feature Continuity Module employs the Riemannian Metric on the Symmetric Positive Definite (SPD) manifold. By measuring the distance of feature statistical distributions in the Riemannian space, it promotes the smooth evolution and continuity of micro-textures across multi-views in a statistical sense. Under the macro-micro synergistic optimization of these two modules, our model can simultaneously improve macro-structural consistency and micro-detail continuity.

Problem

• text-to-3D generation
• macro-topological inconsistency
• micro-geometric discontinuity
• Score Distillation Sampling
• view bias

Innovation

• Manifold-Order Consistency
• Semantic View-Order Constraint
• Riemannian Metric
• Score Distillation Sampling
• Text-to-3D Generation