Theatergen: Character Management with LLM for Consistent Multi-turn Image

[📄Paper]   [🚩Project Page]

Model Architecture

Introduction

We propose Theatergen, a tuning-free method for consistent multi-turn image generation. The key idea is to utilize LLM for character management with layout and id and customize each character to avoid attention leakage. We further propose the CMIGBench for evaluating the consistency in multi-turn image generation.

TODO

• Deployment with GPT interface
• Release Benchmark
• Release code

🔥 News

• [2024.04.26] We have released our code and benchmark

Setup

🔧 Requirements

To install requirements:

pip install -r requirements.txt

🚀 Generate

Generate with CMIGBench or replace with your own demo

python generate.py --task story --sd_version '1.5' --dataset_path CMIGBench

👀 Contact Us

If you have any questions, please feel free to email us at howe4884@outlook.com.

💡Acknowledgement

Our work is based on stable diffusion, Grounded-SAM, T2I-Adapter, and IP-Adapter. We appreciate their outstanding contributions.