提出新方法实现高效且稳定的分数匹配训练，支持高分辨率图像生成。
Local Curvature Smoothing with Stein's Identity for Efficient Score Matching
- 利用Stein恒等式避免计算雅可比迹，提升训练效率。
- 在FID、Inception分数等指标上优于现有方法，媲美主流去噪分数匹配。
- 支持1024×1024高分辨率图像生成，适合高质量图像合成任务。
基于分数的扩散模型（SDMs）的训练依赖于分数匹配，其挑战在于需计算昂贵的雅可比迹。尽管已有多种方法避免该计算，但普遍存在训练不稳或将学习近似为去噪向量场而非真实分数的问题。本文提出一种新型分数匹配变体——局部曲率平滑与Stein恒等式结合（LCSS），通过应用Stein恒等式规避雅可比迹计算，兼具正则化效果与高效性。实验表明，LCSS在样本生成性能上超越现有方法，并在FID、Inception得分和每维比特数等指标上达到与广泛采用的去噪分数匹配相当的水平。此外，LCSS可在1024×1024高分辨率下实现逼真图像生成。
原文摘要 · Abstract (English)
The training of score-based diffusion models (SDMs) is based on score matching. The challenge of score matching is that it includes a computationally expensive Jacobian trace. While several methods have been proposed to avoid this computation, each has drawbacks, such as instability during training and approximating the learning as learning a denoising vector field rather than a true score. We propose a novel score matching variant, local curvature smoothing with Stein's identity (LCSS). The LCSS bypasses the Jacobian trace by applying Stein's identity, enabling regularization effectiveness and efficient computation. We show that LCSS surpasses existing methods in sample generation performance and matches the performance of denoising score matching, widely adopted by most SDMs, in evaluations such as FID, Inception score, and bits per dimension. Furthermore, we show that LCSS enables realistic image generation even at a high resolution of $1024 \times 1024$.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。