PARASOL: Parametric Style Control for Diffusion Image Synthesis

TL;DR

PARASOL is a diffusion model that allows detailed, disentangled control over image content and style, enabling nuanced image creation and stylization through multi-modal conditioning and auxiliary search techniques.

Contribution

It introduces a novel multi-modal diffusion model with parametric style control, trained with specific losses and adapted guidance for independent content and style manipulation.

Findings

Effective disentangled control over content and style.

Enhanced image stylization and search capabilities.

Potential for nuanced, user-driven image generation.

Abstract

We propose PARASOL, a multi-modal synthesis model that enables disentangled, parametric control of the visual style of the image by jointly conditioning synthesis on both content and a fine-grained visual style embedding. We train a latent diffusion model (LDM) using specific losses for each modality and adapt the classifier-free guidance for encouraging disentangled control over independent content and style modalities at inference time. We leverage auxiliary semantic and style-based search to create training triplets for supervision of the LDM, ensuring complementarity of content and style cues. PARASOL shows promise for enabling nuanced control over visual style in diffusion models for image creation and stylization, as well as generative search where text-based search results may be adapted to more closely match user intent by interpolating both content and style descriptors.