Stable Score Distillation for High-Quality 3D Generation

TL;DR

This paper analyzes Score Distillation Sampling (SDS) for 3D generation, decomposes it into key components, identifies its limitations, and proposes Stable Score Distillation (SSD) to improve 3D content quality.

Contribution

The paper offers a detailed analysis of SDS, introduces a new variance-reducing term, and proposes SSD for enhanced 3D generation quality and stability.

Findings

SSD produces higher-fidelity 3D content.

It effectively reduces over-smoothness and implausibility.

The approach is compatible with various frameworks.

Abstract

Although Score Distillation Sampling (SDS) has exhibited remarkable performance in conditional 3D content generation, a comprehensive understanding of its formulation is still lacking, hindering the development of 3D generation. In this work, we decompose SDS as a combination of three functional components, namely mode-seeking, mode-disengaging and variance-reducing terms, analyzing the properties of each. We show that problems such as over-smoothness and implausibility result from the intrinsic deficiency of the first two terms and propose a more advanced variance-reducing term than that introduced by SDS. Based on the analysis, we propose a simple yet effective approach named Stable Score Distillation (SSD) which strategically orchestrates each term for high-quality 3D generation and can be readily incorporated to various 3D generation frameworks and 3D representations. Extensive experiments validate the efficacy of our approach, demonstrating its ability to generate high-fidelity 3D content without succumbing to issues such as over-smoothness.