ART: Anonymous Region Transformer for Variable Multi-Layer Transparent Image Generation
Yifan Pu, Yiming Zhao, Zhicong Tang, Ruihong Yin, Haoxing Ye, Yuhui Yuan, Dong Chen, Jianmin Bao, Sirui Zhang, Yanbin Wang, Lin Liang, Lijuan Wang, Ji Li, Xiu Li, Zhouhui Lian, Gao Huang, Baining Guo
2025-02-26
Summary
This paper talks about ART, a new tool that makes it easier and faster to create multi-layer transparent images, where each layer can be separately edited or customized.
What's the problem?
Creating multi-layer images, where different parts of an image are separated into layers for easier editing, is slow and complicated with current methods. These methods also struggle to handle many layers efficiently and often require detailed instructions for each layer, which can be time-consuming.
What's the solution?
The researchers developed ART (Anonymous Region Transformer), which uses a new approach called 'anonymous region layout' to let the AI decide how to organize image layers based on a general text description. This method reduces the time and computing power needed to generate images with many layers while keeping them consistent and easy to edit. ART also includes a multi-layer autoencoder that ensures high-quality results by encoding and decoding the transparency of each layer.
Why it matters?
This matters because it makes creating complex, editable images much faster and more accessible for designers, artists, and anyone working with digital content. By improving efficiency and quality, ART could transform industries like graphic design, animation, and game development, allowing for more creative freedom and productivity.
Abstract
Multi-layer image generation is a fundamental task that enables users to isolate, select, and edit specific image layers, thereby revolutionizing interactions with generative models. In this paper, we introduce the Anonymous Region Transformer (ART), which facilitates the direct generation of variable multi-layer transparent images based on a global text prompt and an anonymous region layout. Inspired by Schema theory suggests that knowledge is organized in frameworks (schemas) that enable people to interpret and learn from new information by linking it to prior knowledge.}, this anonymous region layout allows the generative model to autonomously determine which set of visual tokens should align with which text tokens, which is in contrast to the previously dominant semantic layout for the image generation task. In addition, the layer-wise region crop mechanism, which only selects the visual tokens belonging to each anonymous region, significantly reduces attention computation costs and enables the efficient generation of images with numerous distinct layers (e.g., 50+). When compared to the full attention approach, our method is over 12 times faster and exhibits fewer layer conflicts. Furthermore, we propose a high-quality multi-layer transparent image autoencoder that supports the direct encoding and decoding of the transparency of variable multi-layer images in a joint manner. By enabling precise control and scalable layer generation, ART establishes a new paradigm for interactive content creation.