CoRe^2: Collect, Reflect and Refine to Generate Better and Faster
Shitong Shao, Zikai Zhou, Dian Xie, Yuetong Fang, Tian Ye, Lichen Bai, Zeke Xie
2025-03-14
Summary
This paper talks about CoRe², a smart add-on for AI image generators that makes them faster and better at creating pictures from text by learning from their own mistakes and refining the process.
What's the problem?
Current AI image tools either make high-quality images slowly or fast images with flaws, and most methods only work for one type of AI model, not all.
What's the solution?
CoRe² uses a three-step method: it collects data on how the AI works, trains a simpler model to handle easy parts faster, then refines the output using a stronger model to fix details.
Why it matters?
This lets AI tools generate high-quality images quickly for things like game design or social media, while working with different types of AI models like Stable Diffusion or LlamaGen.
Abstract
Making text-to-image (T2I) generative model sample both fast and well represents a promising research direction. Previous studies have typically focused on either enhancing the visual quality of synthesized images at the expense of sampling efficiency or dramatically accelerating sampling without improving the base model's generative capacity. Moreover, nearly all inference methods have not been able to ensure stable performance simultaneously on both diffusion models (DMs) and visual autoregressive models (ARMs). In this paper, we introduce a novel plug-and-play inference paradigm, CoRe^2, which comprises three subprocesses: Collect, Reflect, and Refine. CoRe^2 first collects classifier-free guidance (CFG) trajectories, and then use collected data to train a weak model that reflects the easy-to-learn contents while reducing number of function evaluations during inference by half. Subsequently, CoRe^2 employs weak-to-strong guidance to refine the conditional output, thereby improving the model's capacity to generate high-frequency and realistic content, which is difficult for the base model to capture. To the best of our knowledge, CoRe^2 is the first to demonstrate both efficiency and effectiveness across a wide range of DMs, including SDXL, SD3.5, and FLUX, as well as ARMs like LlamaGen. It has exhibited significant performance improvements on HPD v2, Pick-of-Pic, Drawbench, GenEval, and T2I-Compbench. Furthermore, CoRe^2 can be seamlessly integrated with the state-of-the-art Z-Sampling, outperforming it by 0.3 and 0.16 on PickScore and AES, while achieving 5.64s time saving using SD3.5.Code is released at https://github.com/xie-lab-ml/CoRe/tree/main.