Skip to yearly menu bar Skip to main content


DEADiff: An Efficient Stylization Diffusion Model with Disentangled Representations

Tianhao Qi · Shancheng Fang · Yanze Wu · Hongtao Xie · Jiawei Liu · Lang chen · Qian HE · Yongdong Zhang

Arch 4A-E Poster #377
award Highlight
[ ] [ Project Page ]
Wed 19 Jun 5 p.m. PDT — 6:30 p.m. PDT

Abstract: The diffusion-based text-to-image model harbors immense potential in transferring reference style. However, current encoder-based approaches significantly impair the text controllability of text-to-image models while transferring styles. In this paper, we introduce $\textit{DEADiff}$ to address this issue using the following two strategies: 1) a mechanism to decouple the style and semantics of reference images. The decoupled feature representations are first extracted by Q-Formers which are instructed by different text descriptions. Then they are injected into mutually exclusive subsets of cross-attention layers for better disentanglement. 2) A non-reconstructive learning method. The Q-Formers are trained using paired images rather than the identical target, in which the reference image and the ground-truth image are with the same style or semantics. We show that DEADiff attains the best visual stylization results and optimal balance between the text controllability inherent in the text-to-image model and style similarity to the reference image, as demonstrated both quantitatively and qualitatively.

Live content is unavailable. Log in and register to view live content