Frequency-Controlled Diffusion Model for Versatile Text-Guided Image-to-Image Translation

TL;DR

The paper introduces FCDiffusion, a frequency-controlled diffusion framework for versatile text-guided image-to-image translation, leveraging DCT-based filtering to control various image attributes in a unified manner.

Contribution

It proposes a novel frequency-domain filtering approach within diffusion models for flexible, multi-attribute text-guided image translation, enabling diverse applications with a single framework.

Findings

Effective control over style, structure, and layout in image translation.

Outperforms existing methods in qualitative and quantitative evaluations.

Versatile application across multiple image translation tasks.

Abstract

Recently, large-scale text-to-image (T2I) diffusion models have emerged as a powerful tool for image-to-image translation (I2I), allowing open-domain image translation via user-provided text prompts. This paper proposes frequency-controlled diffusion model (FCDiffusion), an end-to-end diffusion-based framework that contributes a novel solution to text-guided I2I from a frequency-domain perspective. At the heart of our framework is a feature-space frequency-domain filtering module based on Discrete Cosine Transform, which filters the latent features of the source image in the DCT domain, yielding filtered image features bearing different DCT spectral bands as different control signals to the pre-trained Latent Diffusion Model. We reveal that control signals of different DCT spectral bands bridge the source image and the T2I generated image in different correlations (e.g., style, structure, layout, contour, etc.), and thus enable versatile I2I applications emphasizing different I2I correlations, including style-guided content creation, image semantic manipulation, image scene translation, and image style translation. Different from related approaches, FCDiffusion establishes a unified text-guided I2I framework suitable for diverse image translation tasks simply by switching among different frequency control branches at inference time. The effectiveness and superiority of our method for text-guided I2I are demonstrated with extensive experiments both qualitatively and quantitatively. Our project is publicly available at: https://xianggao1102.github.io/FCDiffusion/.