High-Resolution Image Synthesis with Latent Diffusion Models

计算机科学 人工智能 推论 修补 像素 灵活性(工程) 计算机视觉 图像(数学) 数学 统计
作者
Robin Rombach,Andreas Blattmann,Dominik Lorenz,Patrick Esser,Björn Ommer
出处
期刊:Cornell University - arXiv 被引量:124
标识
DOI:10.48550/arxiv.2112.10752
摘要

By decomposing the image formation process into a sequential application of denoising autoencoders, diffusion models (DMs) achieve state-of-the-art synthesis results on image data and beyond. Additionally, their formulation allows for a guiding mechanism to control the image generation process without retraining. However, since these models typically operate directly in pixel space, optimization of powerful DMs often consumes hundreds of GPU days and inference is expensive due to sequential evaluations. To enable DM training on limited computational resources while retaining their quality and flexibility, we apply them in the latent space of powerful pretrained autoencoders. In contrast to previous work, training diffusion models on such a representation allows for the first time to reach a near-optimal point between complexity reduction and detail preservation, greatly boosting visual fidelity. By introducing cross-attention layers into the model architecture, we turn diffusion models into powerful and flexible generators for general conditioning inputs such as text or bounding boxes and high-resolution synthesis becomes possible in a convolutional manner. Our latent diffusion models (LDMs) achieve a new state of the art for image inpainting and highly competitive performance on various tasks, including unconditional image generation, semantic scene synthesis, and super-resolution, while significantly reducing computational requirements compared to pixel-based DMs. Code is available at https://github.com/CompVis/latent-diffusion .
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
cx发布了新的文献求助10
1秒前
英俊的铭应助科研通管家采纳,获得10
2秒前
丘比特应助科研通管家采纳,获得10
2秒前
情怀应助科研通管家采纳,获得30
2秒前
大个应助科研通管家采纳,获得10
2秒前
3秒前
CodeCraft应助科研通管家采纳,获得10
3秒前
我是老大应助科研通管家采纳,获得10
3秒前
大个应助科研通管家采纳,获得10
3秒前
烟花应助科研通管家采纳,获得10
3秒前
搜集达人应助胖胖采纳,获得10
3秒前
桐桐应助科研通管家采纳,获得10
3秒前
小赵发布了新的文献求助10
3秒前
科研通AI5应助科研通管家采纳,获得10
3秒前
科研通AI5应助科研通管家采纳,获得10
3秒前
3秒前
3秒前
含蓄怡发布了新的文献求助10
4秒前
4秒前
通行证应助wwss采纳,获得20
5秒前
清欢完成签到 ,获得积分10
6秒前
卡卡西应助澍澍采纳,获得20
6秒前
9秒前
10秒前
Pytong发布了新的文献求助30
10秒前
思源应助易川采纳,获得10
13秒前
四夕水窖完成签到,获得积分10
14秒前
含蓄的荔枝完成签到 ,获得积分10
16秒前
HH完成签到,获得积分10
17秒前
Zyl完成签到 ,获得积分10
19秒前
洁净奄完成签到,获得积分10
22秒前
杨氏完成签到,获得积分10
22秒前
小赵完成签到,获得积分10
28秒前
28秒前
31秒前
wgm关闭了wgm文献求助
31秒前
34秒前
35秒前
justsoso完成签到 ,获得积分10
36秒前
千寻完成签到 ,获得积分10
36秒前
高分求助中
Encyclopedia of Mathematical Physics 2nd edition 888
Introduction to Strong Mixing Conditions Volumes 1-3 500
Tip60 complex regulates eggshell formation and oviposition in the white-backed planthopper, providing effective targets for pest control 400
Optical and electric properties of monocrystalline synthetic diamond irradiated by neutrons 320
共融服務學習指南 300
Essentials of Pharmacoeconomics: Health Economics and Outcomes Research 3rd Edition. by Karen Rascati 300
Peking Blues // Liao San 300
热门求助领域 (近24小时)
化学 材料科学 医学 生物 工程类 有机化学 物理 生物化学 纳米技术 计算机科学 化学工程 内科学 复合材料 物理化学 电极 遗传学 量子力学 基因 冶金 催化作用
热门帖子
关注 科研通微信公众号,转发送积分 3802551
求助须知:如何正确求助?哪些是违规求助? 3348202
关于积分的说明 10337121
捐赠科研通 3064142
什么是DOI,文献DOI怎么找? 1682405
邀请新用户注册赠送积分活动 808168
科研通“疑难数据库(出版商)”最低求助积分说明 763997