arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2107.00630cs.LGstat.ML
收藏

变分扩散模型

Variational Diffusion Models

Google Research(谷歌研究院)

Diederik P. Kingma, Tim Salimans, Ben Poole, Jonathan Ho

AI总结:

本文提出变分扩散模型,通过联合优化噪声调度并利用信噪比简化变分下界,在图像密度估计上超越自回归模型,并实现接近理论最优的无损压缩。

摘要:

基于扩散的生成模型已经展示了在感知上令人印象深刻的合成能力,但它们能否同时成为优秀的基于似然的模型?我们对此给出肯定的回答,并提出一类基于扩散的生成模型,在标准图像密度估计基准上取得了最先进的似然结果。与其他基于扩散的模型不同,我们的方法允许对噪声调度与模型其余部分进行联合高效优化。我们证明变分下界(VLB)可以简化为一个以扩散数据的信噪比表示的非常简短的表达式,从而加深了对这一类模型的理论理解。利用这一见解,我们证明了文献中提出的若干模型之间的等价性。此外,我们证明连续时间 VLB 对噪声调度具有不变性,仅在其端点的信噪比处例外。这使我们能够学习一个使所得 VLB 估计量方差最小化的噪声调度,从而实现更快的优化。将这些进展与架构改进相结合,我们在图像密度估计基准上取得了最先进的似然结果,超越了多年来主导这些基准的自回归模型,并且优化速度通常显著更快。此外,我们展示了如何将该模型用作 bits-back 压缩方案的一部分,并证明了接近理论最优的无损压缩率。代码可在 https://github.com/google-research/vdm 获取。

英文摘要:

Diffusion-based generative models have demonstrated a capacity for perceptually impressive synthesis, but can they also be great likelihood-based models? We answer this in the affirmative, and introduce a family of diffusion-based generative models that obtain state-of-the-art likelihoods on standard image density estimation benchmarks. Unlike other diffusion-based models, our method allows for efficient optimization of the noise schedule jointly with the rest of the model. We show that the variational lower bound (VLB) simplifies to a remarkably short expression in terms of the signal-to-noise ratio of the diffused data, thereby improving our theoretical understanding of this model class. Using this insight, we prove an equivalence between several models proposed in the literature. In addition, we show that the continuous-time VLB is invariant to the noise schedule, except for the signal-to-noise ratio at its endpoints. This enables us to learn a noise schedule that minimizes the variance of the resulting VLB estimator, leading to faster optimization. Combining these advances with architectural improvements, we obtain state-of-the-art likelihoods on image density estimation benchmarks, outperforming autoregressive models that have dominated these benchmarks for many years, with often significantly faster optimization. In addition, we show how to use the model as part of a bits-back compression scheme, and demonstrate lossless compression rates close to the theoretical optimum. Code is available at https://github.com/google-research/vdm .

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑