{"id":623572,"date":"2024-10-22T16:09:09","date_gmt":"2024-10-22T20:09:09","guid":{"rendered":"https:\/\/www.rochester.edu\/newscenter\/?p=623572"},"modified":"2024-10-25T17:42:36","modified_gmt":"2024-10-25T21:42:36","slug":"video-deepfakes-ai-meaning-definition-technology-623572","status":"publish","type":"post","link":"https:\/\/www.rochester.edu\/newscenter\/video-deepfakes-ai-meaning-definition-technology-623572\/","title":{"rendered":"Are video deepfakes powerful enough to influence political discourse?"},"content":{"rendered":"

An expert in AI video generation discusses the technology\u2019s rapid advances\u2014and its current limitations.<\/h2>\n

This presidential cycle has already seen several high-profile examples of people using deepfakes to try to influence voters. <\/span>Deepfakes are images, audio recordings, or videos generated or modified using artificial intelligence (AI) models to depict real or fictional people. <\/span>Recent deepfake examples include manipulated audio of Joe Biden urging voters to stay home during primaries<\/a> and fabricated images of Taylor Swift endorsing Donald Trump<\/a>.<\/p>\n

It appears generative artificial intelligence is an increasingly prominent tool in the misinformation toolbox. Should voters be concerned about being bombarded with phony videos of politicians created with generative AI? An expert in computer vision and deep learning at the University of Rochester<\/a> says that while the technology is rapidly advancing, deepfake video generation remains harder for bad actors to leverage due to its complex nature.<\/p>\n

While OpenAI\u2019s products, including ChatGPT for text generation and DALL-E 3 for image generation, are taking off in popularity, the company has yet to release an equivalent for video generation. According to Chenliang Xu<\/a>, an associate professor of computer science<\/a> at Rochester, the company has released previews of its Sora video generation software but has yet to release the product, which is still undergoing testing and refinement.<\/p>\n

\u201cGenerating video using AI is still an ongoing research topic and a hard problem because it\u2019s what we call multimodal content,\u201d says Xu. \u201cGenerating moving videos along with corresponding audio are difficult problems on their own\u2014and aligning them is even harder.\u201d<\/p>\n

Xu says that his research group was among the first to use artificial neural networks to generate multimodal video in 2017. They started with tasks like providing an image of a violin player and audio of a violin to generate a moving video of a violin player<\/a>. From there, they moved on to problems like generating lip movements<\/a>, and then to creating full talking faces complete with head gestures from a single image<\/a>.<\/p>\n

\u201cNow, we can generate real-time, fully drivable heads and even turn the heads into various styles specified by language descriptions<\/a>,\u201d says Xu.<\/p>\n

\"Diptych
TALKING HEADS:<\/strong> Computer scientist Chenliang Xu and his fellow researchers can generate lifelike talking head videos from an individual photo or even a painting, as demonstrated here with a looping video created from an image of the Mona Lisa and a headshot of Xu. (University of Rochester GIF \/ Luchuan Song)<\/figcaption><\/figure>\n

Challenges with deepfake detection technology<\/strong><\/h3>\n

Xu\u2019s team has also developed technology for deepfake detection<\/a>. He calls it an area that needs extensive further research, noting that it\u2019s easier to build technology to generate deepfakes than to detect them because of the training data needed to build the generalized deepfake detection models.<\/p>\n

\u201c<\/span>Politicians and celebrities are easier to generate than normal people because there is simply more data about them.\u201d<\/div>\n

\u201cIf you want to build a technology that\u2019s able to detect deepfakes, you need to create a database that identifies what are fake images and what are real images,\u201d says Xu. \u201cThat labeling requires an additional layer of human involvement that generation does not.\u201d<\/p>\n

Another concern, he adds, is making a detector that is generalizable to different types of deepfake generators. \u201cYou can make a model that performs well against the techniques you know about, but if someone uses a different model, your detection algorithm will have a hard time capturing that,\u201d he says.<\/p>\n

The easiest targets for video deepfakes<\/strong><\/h3>\n

Having access to good training data is crucial for creating effective generative AI models. As a result, Xu says politicians and celebrities will be the earliest and easiest targets when video generators become widely available.<\/p>\n

\u201cPoliticians and celebrities are easier to generate than normal people because there is simply more data about them,\u201d says Xu. \u201cBecause so much video of them already exists, these models can use it to learn the expressions they show in different situations, along with their voices, their hair, movements, and emotions.\u201d<\/p>\n

But he expects that, at least initially, the training data the \u201cceleb deepfakes\u201d in particular are built on may make them more easily noticeable.<\/p>\n

\u201cIf you used only high-quality photographs to train a model, it will produce similar results,\u201d says Xu. \u201cIt may result in an overly smooth style that you can pick out as a cue to tell it\u2019s a deepfake.\u201d<\/p>\n

Other cues can include how natural a person\u2019s reaction seems, whether they can move their heads, and even the number of teeth shown. But image generators have overcome similar early tells\u2014such as creating hands with six fingers<\/a>\u2014and Xu says enough training data can mitigate these limitations.<\/p>\n

He calls on the research community to invest more effort into developing deepfake detection strategies and grappling with the ethical concerns surrounding the development of these technologies.<\/p>\n

\u201cGenerative models are a tool that in the hands of good people can do good things, but in the hands of bad people can do bad things,\u201d says Xu. \u201cThe technology itself isn\u2019t good or bad, but we need to discuss how to prevent these powerful tools from ending up in the wrong hands and used maliciously.\u201d<\/p>\n","protected":false},"excerpt":{"rendered":"

An expert in AI video generation discusses the technology\u2019s rapid advances\u2014and its current limitations.<\/p>\n","protected":false},"author":1242,"featured_media":623982,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[116],"tags":[24292,18802,18632],"class_list":["post-623572","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-sci-tech","tag-artificial-intelligence","tag-department-of-computer-science","tag-hajim-school-of-engineering-and-applied-sciences"],"acf":[],"yoast_head":"\nAre video deepfakes powerful enough to influence political discourse?<\/title>\n<meta name=\"description\" content=\"A University of Rochester expert in AI video generation and deepfakes discusses the technology\u2019s rapid advances\u2014and its current limitations.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.rochester.edu\/newscenter\/video-deepfakes-ai-meaning-definition-technology-623572\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Are video deepfakes powerful enough to influence political discourse?\" \/>\n<meta property=\"og:description\" content=\"A University of Rochester expert in AI video generation and deepfakes discusses the technology\u2019s rapid advances\u2014and its current limitations.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.rochester.edu\/newscenter\/video-deepfakes-ai-meaning-definition-technology-623572\/\" \/>\n<meta property=\"og:site_name\" content=\"News Center\" \/>\n<meta property=\"article:published_time\" content=\"2024-10-22T20:09:09+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2024-10-25T21:42:36+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/www.rochester.edu\/newscenter\/wp-content\/uploads\/2024\/10\/fea-deepfakes-ai-examples-trump-biden.gif\" \/>\n\t<meta property=\"og:image:width\" content=\"1000\" \/>\n\t<meta property=\"og:image:height\" content=\"600\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/gif\" \/>\n<meta name=\"author\" content=\"Luke Auburn\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Luke Auburn\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"5 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\\\/\\\/www.rochester.edu\\\/newscenter\\\/video-deepfakes-ai-meaning-definition-technology-623572\\\/#article\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/www.rochester.edu\\\/newscenter\\\/video-deepfakes-ai-meaning-definition-technology-623572\\\/\"},\"author\":{\"name\":\"Luke Auburn\",\"@id\":\"https:\\\/\\\/www.rochester.edu\\\/newscenter\\\/#\\\/schema\\\/person\\\/e928dc2863b53a89ece6d40c7992a4e1\"},\"headline\":\"Are video deepfakes powerful enough to influence political discourse?\",\"datePublished\":\"2024-10-22T20:09:09+00:00\",\"dateModified\":\"2024-10-25T21:42:36+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\\\/\\\/www.rochester.edu\\\/newscenter\\\/video-deepfakes-ai-meaning-definition-technology-623572\\\/\"},\"wordCount\":871,\"image\":{\"@id\":\"https:\\\/\\\/www.rochester.edu\\\/newscenter\\\/video-deepfakes-ai-meaning-definition-technology-623572\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/www.rochester.edu\\\/newscenter\\\/wp-content\\\/uploads\\\/2024\\\/10\\\/fea-deepfakes-ai-examples-trump-biden.gif\",\"keywords\":[\"artificial intelligence\",\"Department of Computer Science\",\"Hajim School of Engineering and Applied Sciences\"],\"articleSection\":[\"Science & Technology\"],\"inLanguage\":\"en-US\"},{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/www.rochester.edu\\\/newscenter\\\/video-deepfakes-ai-meaning-definition-technology-623572\\\/\",\"url\":\"https:\\\/\\\/www.rochester.edu\\\/newscenter\\\/video-deepfakes-ai-meaning-definition-technology-623572\\\/\",\"name\":\"Are video deepfakes powerful enough to influence political discourse?\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/www.rochester.edu\\\/newscenter\\\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/www.rochester.edu\\\/newscenter\\\/video-deepfakes-ai-meaning-definition-technology-623572\\\/#primaryimage\"},\"image\":{\"@id\":\"https:\\\/\\\/www.rochester.edu\\\/newscenter\\\/video-deepfakes-ai-meaning-definition-technology-623572\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/www.rochester.edu\\\/newscenter\\\/wp-content\\\/uploads\\\/2024\\\/10\\\/fea-deepfakes-ai-examples-trump-biden.gif\",\"datePublished\":\"2024-10-22T20:09:09+00:00\",\"dateModified\":\"2024-10-25T21:42:36+00:00\",\"author\":{\"@id\":\"https:\\\/\\\/www.rochester.edu\\\/newscenter\\\/#\\\/schema\\\/person\\\/e928dc2863b53a89ece6d40c7992a4e1\"},\"description\":\"A University of Rochester expert in AI video generation and deepfakes discusses the technology\u2019s rapid advances\u2014and its current limitations.\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/www.rochester.edu\\\/newscenter\\\/video-deepfakes-ai-meaning-definition-technology-623572\\\/#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/www.rochester.edu\\\/newscenter\\\/video-deepfakes-ai-meaning-definition-technology-623572\\\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/www.rochester.edu\\\/newscenter\\\/video-deepfakes-ai-meaning-definition-technology-623572\\\/#primaryimage\",\"url\":\"https:\\\/\\\/www.rochester.edu\\\/newscenter\\\/wp-content\\\/uploads\\\/2024\\\/10\\\/fea-deepfakes-ai-examples-trump-biden.gif\",\"contentUrl\":\"https:\\\/\\\/www.rochester.edu\\\/newscenter\\\/wp-content\\\/uploads\\\/2024\\\/10\\\/fea-deepfakes-ai-examples-trump-biden.gif\",\"width\":1000,\"height\":600,\"caption\":\"FACE THE FACTS: Using emotional listener generation technology, University of Rochester computer scientists have generated deepfakes of Joe Biden displaying different emotional expressions in reaction to Donald Trump. (University of Rochester GIF \\\/ Luchuan Song)\"},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/www.rochester.edu\\\/newscenter\\\/video-deepfakes-ai-meaning-definition-technology-623572\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/www.rochester.edu\\\/newscenter\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Are video deepfakes powerful enough to influence political discourse?\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/www.rochester.edu\\\/newscenter\\\/#website\",\"url\":\"https:\\\/\\\/www.rochester.edu\\\/newscenter\\\/\",\"name\":\"News Center\",\"description\":\"University of Rochester\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/www.rochester.edu\\\/newscenter\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"},{\"@type\":\"Person\",\"@id\":\"https:\\\/\\\/www.rochester.edu\\\/newscenter\\\/#\\\/schema\\\/person\\\/e928dc2863b53a89ece6d40c7992a4e1\",\"name\":\"Luke Auburn\",\"url\":\"https:\\\/\\\/www.rochester.edu\\\/newscenter\\\/author\\\/lauburn\\\/\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Are video deepfakes powerful enough to influence political discourse?","description":"A University of Rochester expert in AI video generation and deepfakes discusses the technology\u2019s rapid advances\u2014and its current limitations.","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/www.rochester.edu\/newscenter\/video-deepfakes-ai-meaning-definition-technology-623572\/","og_locale":"en_US","og_type":"article","og_title":"Are video deepfakes powerful enough to influence political discourse?","og_description":"A University of Rochester expert in AI video generation and deepfakes discusses the technology\u2019s rapid advances\u2014and its current limitations.","og_url":"https:\/\/www.rochester.edu\/newscenter\/video-deepfakes-ai-meaning-definition-technology-623572\/","og_site_name":"News Center","article_published_time":"2024-10-22T20:09:09+00:00","article_modified_time":"2024-10-25T21:42:36+00:00","og_image":[{"url":"https:\/\/www.rochester.edu\/newscenter\/wp-content\/uploads\/2024\/10\/fea-deepfakes-ai-examples-trump-biden.gif","width":1000,"height":600,"type":"image\/gif"}],"author":"Luke Auburn","twitter_card":"summary_large_image","twitter_misc":{"Written by":"Luke Auburn","Est. reading time":"5 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/www.rochester.edu\/newscenter\/video-deepfakes-ai-meaning-definition-technology-623572\/#article","isPartOf":{"@id":"https:\/\/www.rochester.edu\/newscenter\/video-deepfakes-ai-meaning-definition-technology-623572\/"},"author":{"name":"Luke Auburn","@id":"https:\/\/www.rochester.edu\/newscenter\/#\/schema\/person\/e928dc2863b53a89ece6d40c7992a4e1"},"headline":"Are video deepfakes powerful enough to influence political discourse?","datePublished":"2024-10-22T20:09:09+00:00","dateModified":"2024-10-25T21:42:36+00:00","mainEntityOfPage":{"@id":"https:\/\/www.rochester.edu\/newscenter\/video-deepfakes-ai-meaning-definition-technology-623572\/"},"wordCount":871,"image":{"@id":"https:\/\/www.rochester.edu\/newscenter\/video-deepfakes-ai-meaning-definition-technology-623572\/#primaryimage"},"thumbnailUrl":"https:\/\/www.rochester.edu\/newscenter\/wp-content\/uploads\/2024\/10\/fea-deepfakes-ai-examples-trump-biden.gif","keywords":["artificial intelligence","Department of Computer Science","Hajim School of Engineering and Applied Sciences"],"articleSection":["Science & Technology"],"inLanguage":"en-US"},{"@type":"WebPage","@id":"https:\/\/www.rochester.edu\/newscenter\/video-deepfakes-ai-meaning-definition-technology-623572\/","url":"https:\/\/www.rochester.edu\/newscenter\/video-deepfakes-ai-meaning-definition-technology-623572\/","name":"Are video deepfakes powerful enough to influence political discourse?","isPartOf":{"@id":"https:\/\/www.rochester.edu\/newscenter\/#website"},"primaryImageOfPage":{"@id":"https:\/\/www.rochester.edu\/newscenter\/video-deepfakes-ai-meaning-definition-technology-623572\/#primaryimage"},"image":{"@id":"https:\/\/www.rochester.edu\/newscenter\/video-deepfakes-ai-meaning-definition-technology-623572\/#primaryimage"},"thumbnailUrl":"https:\/\/www.rochester.edu\/newscenter\/wp-content\/uploads\/2024\/10\/fea-deepfakes-ai-examples-trump-biden.gif","datePublished":"2024-10-22T20:09:09+00:00","dateModified":"2024-10-25T21:42:36+00:00","author":{"@id":"https:\/\/www.rochester.edu\/newscenter\/#\/schema\/person\/e928dc2863b53a89ece6d40c7992a4e1"},"description":"A University of Rochester expert in AI video generation and deepfakes discusses the technology\u2019s rapid advances\u2014and its current limitations.","breadcrumb":{"@id":"https:\/\/www.rochester.edu\/newscenter\/video-deepfakes-ai-meaning-definition-technology-623572\/#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.rochester.edu\/newscenter\/video-deepfakes-ai-meaning-definition-technology-623572\/"]}]},{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/www.rochester.edu\/newscenter\/video-deepfakes-ai-meaning-definition-technology-623572\/#primaryimage","url":"https:\/\/www.rochester.edu\/newscenter\/wp-content\/uploads\/2024\/10\/fea-deepfakes-ai-examples-trump-biden.gif","contentUrl":"https:\/\/www.rochester.edu\/newscenter\/wp-content\/uploads\/2024\/10\/fea-deepfakes-ai-examples-trump-biden.gif","width":1000,"height":600,"caption":"FACE THE FACTS: Using emotional listener generation technology, University of Rochester computer scientists have generated deepfakes of Joe Biden displaying different emotional expressions in reaction to Donald Trump. (University of Rochester GIF \/ Luchuan Song)"},{"@type":"BreadcrumbList","@id":"https:\/\/www.rochester.edu\/newscenter\/video-deepfakes-ai-meaning-definition-technology-623572\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/www.rochester.edu\/newscenter\/"},{"@type":"ListItem","position":2,"name":"Are video deepfakes powerful enough to influence political discourse?"}]},{"@type":"WebSite","@id":"https:\/\/www.rochester.edu\/newscenter\/#website","url":"https:\/\/www.rochester.edu\/newscenter\/","name":"News Center","description":"University of Rochester","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.rochester.edu\/newscenter\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Person","@id":"https:\/\/www.rochester.edu\/newscenter\/#\/schema\/person\/e928dc2863b53a89ece6d40c7992a4e1","name":"Luke Auburn","url":"https:\/\/www.rochester.edu\/newscenter\/author\/lauburn\/"}]}},"_links":{"self":[{"href":"https:\/\/www.rochester.edu\/newscenter\/wp-json\/wp\/v2\/posts\/623572","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.rochester.edu\/newscenter\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.rochester.edu\/newscenter\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.rochester.edu\/newscenter\/wp-json\/wp\/v2\/users\/1242"}],"replies":[{"embeddable":true,"href":"https:\/\/www.rochester.edu\/newscenter\/wp-json\/wp\/v2\/comments?post=623572"}],"version-history":[{"count":12,"href":"https:\/\/www.rochester.edu\/newscenter\/wp-json\/wp\/v2\/posts\/623572\/revisions"}],"predecessor-version":[{"id":625012,"href":"https:\/\/www.rochester.edu\/newscenter\/wp-json\/wp\/v2\/posts\/623572\/revisions\/625012"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.rochester.edu\/newscenter\/wp-json\/wp\/v2\/media\/623982"}],"wp:attachment":[{"href":"https:\/\/www.rochester.edu\/newscenter\/wp-json\/wp\/v2\/media?parent=623572"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.rochester.edu\/newscenter\/wp-json\/wp\/v2\/categories?post=623572"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.rochester.edu\/newscenter\/wp-json\/wp\/v2\/tags?post=623572"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}