Applications of Generative AI to Media

Q3 Engineering
Brent Rabowsky
{"title":"Applications of Generative AI to Media","authors":"Brent Rabowsky","doi":"10.5594/JMI.2023.3297238","DOIUrl":null,"url":null,"abstract":"When ChatGPT was released in November 2022, the public became aware of a new era of artificial intelligence (AI) built on a class of machine learning models known as foundation models (FMs). Although FMs burst upon the scene seemingly without precedent, the roots of FMs go back several years earlier. A key distinguishing factor of FMs versus older models is their powerful applications to Generative AI (GenAI), the use of models to generate new content and transform existing content. GenAI models tend to be general-purpose and more flexible than previous kinds of AI models, which are limited to specific predictive tasks such as object detection in images, text classification, forecasting, or detecting data anomalies. Although there were some other kinds of GenAI models before the rise of FMs, FMs are much more powerful and are an enormous leap over their predecessors.","PeriodicalId":49512,"journal":{"name":"SMPTE Motion Imaging Journal","volume":"132 8","pages":"53-57"},"PeriodicalIF":0.0000,"publicationDate":"2023-09-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"SMPTE Motion Imaging Journal","FirstCategoryId":"1085","ListUrlMain":"https://ieeexplore.ieee.org/document/10243436/","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"Engineering","Score":null,"Total":0}
引用次数: 0

Abstract

When ChatGPT was released in November 2022, the public became aware of a new era of artificial intelligence (AI) built on a class of machine learning models known as foundation models (FMs). Although FMs burst upon the scene seemingly without precedent, the roots of FMs go back several years earlier. A key distinguishing factor of FMs versus older models is their powerful applications to Generative AI (GenAI), the use of models to generate new content and transform existing content. GenAI models tend to be general-purpose and more flexible than previous kinds of AI models, which are limited to specific predictive tasks such as object detection in images, text classification, forecasting, or detecting data anomalies. Although there were some other kinds of GenAI models before the rise of FMs, FMs are much more powerful and are an enormous leap over their predecessors.
生成式AI在媒体中的应用
当ChatGPT于2022年11月发布时,公众意识到建立在一类被称为基础模型(FMs)的机器学习模型上的人工智能(AI)的新时代。虽然FMs的出现似乎没有先例,但FMs的起源可以追溯到几年前。FMs与旧模型的一个关键区别因素是它们对生成式人工智能(GenAI)的强大应用,即使用模型生成新内容并转换现有内容。GenAI模型往往是通用的,比以前的人工智能模型更灵活,这些模型仅限于特定的预测任务,如图像中的对象检测、文本分类、预测或检测数据异常。尽管在FMs兴起之前已经有一些其他类型的GenAI模型,但FMs要强大得多,并且比它们的前辈有了巨大的飞跃。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 求助全文
来源期刊
SMPTE Motion Imaging Journal
SMPTE Motion Imaging Journal 工程技术-成像科学与照相技术
CiteScore
0.60
自引率
0.00%
发文量
0
审稿时长
>12 weeks
期刊介绍: The SMPTE Motion Imaging Journal is the key publication of the Society, consistently ranked by our members as the most valuable benefit of their SMPTE membership. Each issue of the Journal explores a theme in great depth, with peer-reviewed technical articles from leading academics, researchers and engineers working at the top companies worldwide. You''ll expand your knowledge on topics like image processing, display technologies, audio, compression, standards, digital cinema, distribution and machine learning and much more. For additional coverage of each month''s topic, the Journal features more exclusive articles online.
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
确定
请完成安全验证×
copy
已复制链接
快去分享给好友吧!
我知道了
右上角分享
点击右上角分享
0
联系我们:info@booksci.cn Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术官方微信