Skip to content

PeterYYZhang/LayerCraft

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

7 Commits
 
 
 
 

Repository files navigation

LayerCraft

This is the official repository for "LayerCraft: Enhancing Text-to-Image Generation with CoT Reasoning and Layered Object Integration"

Workflow demonstration for LayerCraft. The user initially provides a simple prompt, “Alice in a wonderland,” and the framework generates an image by employing Chain-of-Thought reasoning to determine both content and spatial arrangements. Subsequently, the user applies a square mask to remove the second mushroom from the left and specifies the addition of a cute lion. After manual region selection, the framework seamlessly integrates the lion into the scene.

Abstract

Text-to-image generation (T2I) has become a key area of research with broad applications. However, existing methods often struggle with complex spatial relationships and fine-grained control over multiple concepts. Many existing approaches require significant architectural modifications, extensive training, or expert-level prompt engineer-ing. To address these challenges, we introduce LayerCraft, an automated framework that leverages large language models (LLMs) as autonomous agents for structured procedural generation. LayerCraft enables users to customize objects within an image and supports narrative-driven creation with minimal effort. At its core, the system includes a coordinator agent that directs the process, along with two specialized agents: ChainArchitect, which employs chain-of-thought (CoT) reasoning to generate a dependency-aware 3D layout for precise instance-level control, and the Object-Integration Network (OIN), which utilizes LoRA fine-tuning on pre-trained T2I models to seamlessly blend objects into specified regions of an image based on textual prompts—without requiring architectural changes. Extensive evaluations demonstrate LayerCraft’s versatility in applications ranging from multi-concept customization to storytelling. By providing non-experts with intuitive, precise control over T2I generation, our framework democratizes creative image creation.

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published