Diffusion-Based Representation Learning

TL;DR

This paper introduces a diffusion-based framework for unsupervised representation learning that leverages a new denoising score matching formulation, enabling control over detail levels and improving semi-supervised image classification.

Contribution

It presents a novel diffusion-based approach for unsupervised representation learning that does not require supervision and allows manual control of encoded detail levels.

Findings

Improved semi-supervised image classification performance.

Comparable or superior representation quality on downstream tasks.

Effective control over the level of detail in learned representations.

Abstract

Diffusion-based methods represented as stochastic differential equations on a continuous-time domain have recently proven successful as a non-adversarial generative model. Training such models relies on denoising score matching, which can be seen as multi-scale denoising autoencoders. Here, we augment the denoising score matching framework to enable representation learning without any supervised signal. GANs and VAEs learn representations by directly transforming latent codes to data samples. In contrast, the introduced diffusion-based representation learning relies on a new formulation of the denoising score matching objective and thus encodes the information needed for denoising. We illustrate how this difference allows for manual control of the level of details encoded in the representation. Using the same approach, we propose to learn an infinite-dimensional latent code that achieves improvements of state-of-the-art models on semi-supervised image classification. We also compare the quality of learned representations of diffusion score matching with other methods like autoencoder and contrastively trained systems through their performances on downstream tasks.