arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2404.00879cs.CV
收藏

扩散模型中与模型无关的人类偏好反演

Model-Agnostic Human Preference Inversion in Diffusion Models

Purdue University(普渡大学)

Jeeyung Kim, Ze Wang, Qiang Qiu

AI总结:

本文提出提示自适应人类偏好反演(PAHI)方法,通过在不微调扩散模型的情况下为每个提示优化先验噪声分布,实现符合人类偏好的高质量一步图像生成,并以极小计算代价显著提升图像质量。

摘要:

由于扩散模型中多步采样带来的高计算成本,高效的文本到图像生成仍然是一项具有挑战性的任务。尽管对预训练扩散模型进行蒸馏已在减少采样步数方面取得了成功,但低步数图像生成在质量上往往仍有所不足。在本研究中,我们提出了一种新颖的采样设计,以实现符合人类偏好的高质量一步图像生成,并特别着重探索先验噪声分布的影响。我们的方法称为提示自适应人类偏好反演(PAHI),它在无需微调扩散模型的情况下,基于人类偏好为每个提示优化噪声分布。我们的实验表明,定制化的噪声分布能够显著提升图像质量,同时仅带来极小的计算成本增加。我们的发现强调了噪声优化的重要性,并为高效且高质量的文本到图像合成铺平了道路。

英文摘要:

Efficient text-to-image generation remains a challenging task due to the high computational costs associated with the multi-step sampling in diffusion models. Although distillation of pre-trained diffusion models has been successful in reducing sampling steps, low-step image generation often falls short in terms of quality. In this study, we propose a novel sampling design to achieve high-quality one-step image generation aligning with human preferences, particularly focusing on exploring the impact of the prior noise distribution. Our approach, Prompt Adaptive Human Preference Inversion (PAHI), optimizes the noise distributions for each prompt based on human preferences without the need for fine-tuning diffusion models. Our experiments showcase that the tailored noise distributions significantly improve image quality with only a marginal increase in computational cost. Our findings underscore the importance of noise optimization and pave the way for efficient and high-quality text-to-image synthesis.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑