AIComposer: Any Style and Content Image Composition via Feature Integration

TL;DR

AIComposer introduces a novel cross-domain image composition method that eliminates the need for text prompts, effectively preserves content, and achieves superior stylization and composition quality using diffusion models and feature integration.

Contribution

It is the first to enable text-prompt-free cross-domain image composition with a simple MLP and local cross-attention, improving robustness and style transfer without additional training.

Findings

Outperforms state-of-the-art in LPIPS and CSD metrics.

Preserves foreground content effectively during stylization.

Demonstrates robustness across diverse styles and contents.

Abstract

Image composition has advanced significantly with large-scale pre-trained T2I diffusion models. Despite progress in same-domain composition, cross-domain composition remains under-explored. The main challenges are the stochastic nature of diffusion models and the style gap between input images, leading to failures and artifacts. Additionally, heavy reliance on text prompts limits practical applications. This paper presents the first cross-domain image composition method that does not require text prompts, allowing natural stylization and seamless compositions. Our method is efficient and robust, preserving the diffusion prior, as it involves minor steps for backward inversion and forward denoising without training the diffuser. Our method also uses a simple multilayer perceptron network to integrate CLIP features from foreground and background, manipulating diffusion with a local cross-attention strategy. It effectively preserves foreground content while enabling stable stylization without a pre-stylization network. Finally, we create a benchmark dataset with diverse contents and styles for fair evaluation, addressing the lack of testing datasets for cross-domain image composition. Our method outperforms state-of-the-art techniques in both qualitative and quantitative evaluations, significantly improving the LPIPS score by 30.5% and the CSD metric by 18.1%. We believe our method will advance future research and applications. Code and benchmark at https://github.com/sherlhw/AIComposer.