Ever wondered what Mona Lisa would look like rapping? Microsoft 開始する，打ち上げるs VASA-1 AI bot that can make images talk - with eerily 現実主義の results

• e-mail

The 境界 between what's real and what's not is becoming ever thinner thanks to a new AI 道具 from Microsoft.?

Called?VASA-1, the 科学（工学）技術?transforms a still image of a person's 直面する into an animated clip of them talking or singing.?

Lip movements are 'exquisitely synchronised' with 音声部の to make it seem like the 支配する has come to life, the tech 巨大（な） （人命などを）奪う，主張するs.?

In one example, Leonardo da Vinci's 16th century masterpiece 'The Mona Lisa' starts rapping crudely in an American accent.?

However, Microsoft 収容する／認めるs the 道具 could be 'misused for impersonating humans' and is not 解放（する）ing it to the public.?

Microsoft's new 道具?VASA-1 can 生成する clips of people talking from a still image and 音声部の of someone talking - but the tech 巨大（な） isn't 解放（する）ing it any time soon

READ MORE?Scientists say 偽の 直面するs created by AI look MORE real than human 直面するs?

AI 'hyperrealism' means 偽の 直面するs are perceived as more real than human ones?

宣伝

VASA-1 takes a static image of a 直面する ? whether it's a photo of a real person or an artwork or 製図／抽選 of someone fictional.?

It then 'meticulously' matches this up with 音声部の of speech 'from any person' to make the 直面する come to life.?

The AI was trained with a library of facial 表現s, which even lets it animate the still image even in real time ? so as the 音声部の is 存在 spoken.?

In a?blog 地位，任命する,?Microsoft 研究員s 述べる VASA as a '枠組み for 生成するing lifelike talking 直面するs of 事実上の characters'.?

'It 覆うs the way for real-time 約束／交戦s with lifelike avatars that emulate human conversational 行為s,' they say.?

'Our method is 有能な of not only producing precious lip-音声部の synchronisation, but also 逮捕（する）ing a large spectrum of emotions and expressive facial nuances and natural 長，率いる 動議s that 与える／捧げる to the perception of realism and liveliness.'?

ーに関して／ーの点でs of use 事例／患者s, the team thinks VASA-1 could enable 数字表示式の AI avatars to 'engage with us in ways that are as natural and intuitive as interactions with real humans'.?

But 専門家s have 株d their 関心s around the 科学（工学）技術, which if 解放（する）d could make people appear to say things that they never said.?

VASA-1 要求するs a static image of a 直面する ? whether it's a photo of a real person or an artwork or 製図／抽選 of someone imaginary. It 'meticulously' matches this up with 音声部の of speech 'from any person' to make the 直面する come to life

Microsoft's team said VASA-1 is 'not ーするつもりであるd to create content that is used to 誤って導く or deceive'

READ MORE I tried an AI 直面する scanning app dubbed the 'most 乱すing 場所／位置 on the internet

PimEyes searches your 直面する using AI and biometrics?

宣伝

Another 可能性のある 危険 is 詐欺, as people online could be duped by a 偽の message from the image of someone they 信用.?

Jake Moore, a 安全 specialist at ESET, said 'seeing is most definitely not believing anymore'.?

'As this 科学（工学）技術 改善するs, it is a race against time to make sure everyone is fully aware of what is 有能な and that they should think twice before they 受託する correspondence as 本物の,' he told MailOnline.?

心配するing 関心s that the public might have, the Microsoft 専門家s said VASA-1 is 'not ーするつもりであるd to create content that is used to 誤って導く or deceive'.?

'However, like other 関係のある content 世代 techniques, it could still 潜在的に be misused for impersonating humans,' they 追加する.?

'We are …に反対するd to any 行為 to create 誤って導くing or harmful contents of real persons, and are 利益／興味d in 適用するing our technique for 前進するing 偽造 （犯罪，病気などの）発見.?

'現在／一般に, the ビデオs 生成するd by this method still 含む／封じ込める identifiable artifacts, and the 数値／数字による 分析 shows that there's still a gap to 達成する the authenticity of real ビデオs.'?

Microsoft 収容する／認めるs that 存在するing techniques are still far from '達成するing the authenticity of natural talking 直面するs', but the 能力 of AI is growing 速く.?

関わりなく the 直面する in the image, the 道具 can form 現実主義の facial 表現s that match the sounds of the words 存在 spoken??

によれば 研究員s at Australian 国家の University, 偽の 直面するs made by AI seem more 現実主義の than human 直面するs.?

These 専門家s 警告するd that AI depictions of people tend to have a 'hyperrealism', with 直面するs that are more in-割合, and people mistake this as a 調印する of humanness.

Another 熟考する／考慮する by 専門家s at Lancaster University 設立する 偽の AI 直面するs appear more 信頼できる, which has 関わりあい／含蓄s for online privacy.??

一方／合間, OpenAI, the creator of the famous ChatGPT bot, introduced its 'terrifying' text-to-ビデオ 道具 Sora?in February, which?can make ultra-現実主義の AI ビデオ clips based 単独で on short, descriptive text 誘発するs.?

This でっちあげる，人を罪に陥れる of an AI-生成するd ビデオ of Tokyo created by OpenAI's Sora shocked 専門家s with its 'terrifying' realism??

In 返答 to the 誘発する 'a cat waking up its sleeping owner 需要・要求するing breakfast', Sora returned this film?

A 献身的な page on OpenAI's website has a rich gallery of the AI-made films, from a man walking on a treadmill to reflections in the windows of a moving train and a cat waking up its owner.

However, 専門家s 警告するd it could wipe out entire 産業s such as film 生産／産物 and lead to a rise in 深い 偽の ビデオs 主要な up to the US 大統領の 選挙.?

'The idea that an AI can create a hyper-現実主義の ビデオ of, say, a 政治家，政治屋 doing something untoward should （犯罪の）一味 alarm bells as we enter into the most 選挙-激しい year in human history,'?said Dr Andrew Rogoyski from the University of Surrey.

A 研究 paper 述べるing Microsoft's new 道具 has been published as a pre-print.?