Stable Score Distillation for High-Quality 3D Generation
This addresses a bottleneck in 3D content generation for applications like graphics and VR, though it is incremental as it builds on existing SDS methods.
The paper tackled the problem of Score Distillation Sampling (SDS) in 3D generation, which suffers from issues like over-smoothness, by proposing Stable Score Distillation (SSD) that strategically orchestrates functional components to generate high-fidelity 3D content without such problems.
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.