Estimating High Order Gradients of the Data Distribution by Denoising

Chenlin Meng, Yang Song, Wenzhe Li, Stefano Ermon

Keywords: score matching, Langevin dynamics, denoising, generative modeling

Abstract: The first order derivative of a data density can be estimated efficiently by denoising score matching, and has become an important component in many applications, such as image generation and audio synthesis. Higher order derivatives provide additional local information about the data distribution and enable new applications. Although they can be estimated via automatic differentiation of a learned density model, this can amplify estimation errors and is expensive in high dimensional settings. To overcome these limitations, we propose a method to directly estimate high order derivatives (scores) of a data density from samples. We first show that denoising score matching can be interpreted as a particular case of Tweedie’s formula. By leveraging Tweedie’s formula on higher order moments, we generalize denoising score matching to estimate higher order derivatives. We demonstrate empirically that models trained with the proposed method can approximate second order derivatives more efficiently and accurately than via automatic differentiation. We show that our models can be used to quantify uncertainty in denoising and to improve the mixing speed of Langevin dynamics via Ozaki discretization for sampling synthetic data and natural images.

Code Of Conduct: I certify that all co-authors of this work have read and commit to adhering to the NeurIPS Statement on Ethics, Fairness, Inclusivity, and Code of Conduct.

TL;DR: We generalize denoising score matching to estimate higher order gradients of the log data density from samples.

Code: https://github.com/chenlin9/high_order_dsm

Community Implementations: [ 1 code implementation](https://www.catalyzex.com/paper/arxiv:2111.04726/code)