Sequential Data Generation with Groupwise Diffusion Process

TL;DR

The paper introduces the Groupwise Diffusion Model (GDM), a novel sequential data generation framework that enhances diffusion models by incorporating group-based diffusion, enabling interpretable latent spaces and applications like image editing and attribute disentanglement.

Contribution

GDM extends diffusion models to a groupwise, sequential process, allowing flexible data generation, interpretability, and frequency domain extensions for hierarchical representations.

Findings

GDM generalizes autoregressive and cascaded diffusion models.

GDM enables interpretable, group-wise latent spaces.

Applications include disentanglement, editing, and variation generation.

Abstract

We present the Groupwise Diffusion Model (GDM), which divides data into multiple groups and diffuses one group at one time interval in the forward diffusion process. GDM generates data sequentially from one group at one time interval, leading to several interesting properties. First, as an extension of diffusion models, GDM generalizes certain forms of autoregressive models and cascaded diffusion models. As a unified framework, GDM allows us to investigate design choices that have been overlooked in previous works, such as data-grouping strategy and order of generation. Furthermore, since one group of the initial noise affects only a certain group of the generated data, latent space now possesses group-wise interpretable meaning. We can further extend GDM to the frequency domain where the forward process sequentially diffuses each group of frequency components. Dividing the frequency bands of the data as groups allows the latent variables to become a hierarchical representation where individual groups encode data at different levels of abstraction. We demonstrate several applications of such representation including disentanglement of semantic attributes, image editing, and generating variations.