Consistent Flow Distillation for Text-to-3D Generation
This addresses a key limitation in 3D generation for applications like graphics and design, though it appears incremental as it builds on existing SDS methods.
The paper tackled the problem of degraded visual quality and diversity in text-to-3D generation using Score Distillation Sampling (SDS) by proposing Consistent Flow Distillation (CFD), which leverages gradient-based sampling and multi-view consistent Gaussian noise to achieve significant improvements over previous methods.
Score Distillation Sampling (SDS) has made significant strides in distilling image-generative models for 3D generation. However, its maximum-likelihood-seeking behavior often leads to degraded visual quality and diversity, limiting its effectiveness in 3D applications. In this work, we propose Consistent Flow Distillation (CFD), which addresses these limitations. We begin by leveraging the gradient of the diffusion ODE or SDE sampling process to guide the 3D generation. From the gradient-based sampling perspective, we find that the consistency of 2D image flows across different viewpoints is important for high-quality 3D generation. To achieve this, we introduce multi-view consistent Gaussian noise on the 3D object, which can be rendered from various viewpoints to compute the flow gradient. Our experiments demonstrate that CFD, through consistent flows, significantly outperforms previous methods in text-to-3D generation.