Saved in:
Bibliographic Details
Main Authors: Chai, Weilong, Zheng, DanDan, Cao, Jiajiong, Chen, Zhiquan, Wang, Changbao, Ma, Chenguang
Format: Preprint
Published: 2023
Subjects:
Online Access:https://arxiv.org/abs/2312.08887
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909331046268928
author Chai, Weilong
Zheng, DanDan
Cao, Jiajiong
Chen, Zhiquan
Wang, Changbao
Ma, Chenguang
author_facet Chai, Weilong
Zheng, DanDan
Cao, Jiajiong
Chen, Zhiquan
Wang, Changbao
Ma, Chenguang
contents Text-to-image diffusion models (SD) exhibit significant advancements while requiring extensive computational resources. Existing acceleration methods usually require extensive training and are not universally applicable. LCM-LoRA, trainable once for diverse models, offers universality but rarely considers ensuring the consistency of generated content before and after acceleration. This paper proposes SpeedUpNet (SUN), an innovative acceleration module, to address the challenges of universality and consistency. Exploiting the role of cross-attention layers in U-Net for SD models, we introduce an adapter specifically designed for these layers, quantifying the offset in image generation caused by negative prompts relative to positive prompts. This learned offset demonstrates stability across a range of models, enhancing SUN's universality. To improve output consistency, we propose a Multi-Step Consistency (MSC) loss, which stabilizes the offset and ensures fidelity in accelerated content. Experiments on SD v1.5 show that SUN leads to an overall speedup of more than 10 times compared to the baseline 25-step DPM-solver++, and offers two extra advantages: (1) training-free integration into various fine-tuned Stable-Diffusion models and (2) state-of-the-art FIDs of the generated data set before and after acceleration guided by random combinations of positive and negative prompts. Code is available: https://williechai.github.io/speedup-plugin-for-stable-diffusions.github.io.
format Preprint
id arxiv_https___arxiv_org_abs_2312_08887
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle SpeedUpNet: A Plug-and-Play Adapter Network for Accelerating Text-to-Image Diffusion Models
Chai, Weilong
Zheng, DanDan
Cao, Jiajiong
Chen, Zhiquan
Wang, Changbao
Ma, Chenguang
Computer Vision and Pattern Recognition
Machine Learning
Text-to-image diffusion models (SD) exhibit significant advancements while requiring extensive computational resources. Existing acceleration methods usually require extensive training and are not universally applicable. LCM-LoRA, trainable once for diverse models, offers universality but rarely considers ensuring the consistency of generated content before and after acceleration. This paper proposes SpeedUpNet (SUN), an innovative acceleration module, to address the challenges of universality and consistency. Exploiting the role of cross-attention layers in U-Net for SD models, we introduce an adapter specifically designed for these layers, quantifying the offset in image generation caused by negative prompts relative to positive prompts. This learned offset demonstrates stability across a range of models, enhancing SUN's universality. To improve output consistency, we propose a Multi-Step Consistency (MSC) loss, which stabilizes the offset and ensures fidelity in accelerated content. Experiments on SD v1.5 show that SUN leads to an overall speedup of more than 10 times compared to the baseline 25-step DPM-solver++, and offers two extra advantages: (1) training-free integration into various fine-tuned Stable-Diffusion models and (2) state-of-the-art FIDs of the generated data set before and after acceleration guided by random combinations of positive and negative prompts. Code is available: https://williechai.github.io/speedup-plugin-for-stable-diffusions.github.io.
title SpeedUpNet: A Plug-and-Play Adapter Network for Accelerating Text-to-Image Diffusion Models
topic Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2312.08887