Diffusion Models For Multi-Modal Generative Modeling

TL;DR

This paper introduces a unified multi-modal diffusion model that can generate and handle various types of data simultaneously, advancing the capabilities of diffusion-based generative modeling.

Contribution

The paper proposes a novel multi-modal diffusion framework with a shared backbone and modality-specific decoders, enabling multi-task learning and multi-modal data generation.

Findings

Effective in image transition and masked-image training

Supports joint image-label and image-representation modeling

Shows promising results on ImageNet

Abstract

Diffusion-based generative modeling has been achieving state-of-the-art results on various generation tasks. Most diffusion models, however, are limited to a single-generation modeling. Can we generalize diffusion models with the ability of multi-modal generative training for more generalizable modeling? In this paper, we propose a principled way to define a diffusion model by constructing a unified multi-modal diffusion model in a common diffusion space. We define the forward diffusion process to be driven by an information aggregation from multiple types of task-data, e.g., images for a generation task and labels for a classification task. In the reverse process, we enforce information sharing by parameterizing a shared backbone denoising network with additional modality-specific decoder heads. Such a structure can simultaneously learn to generate different types of multi-modal data with a multi-task loss, which is derived from a new multi-modal variational lower bound that generalizes the standard diffusion model. We propose several multimodal generation settings to verify our framework, including image transition, masked-image training, joint image-label and joint image-representation generative modeling. Extensive experimental results on ImageNet indicate the effectiveness of our framework for various multi-modal generative modeling, which we believe is an important research direction worthy of more future explorations.