A new artificial intelligence created by researchers at the Massachusetts Institute of Technology pulls off a staggering feat: by analyzing only a short audio clip of a person's voice, it reconstructs what they might look like in real life.
麻省理工學(xué)院的研究人員創(chuàng)造的一種新的人工智能取得了驚人的成就：通過(guò)僅分析一個(gè)人聲音的短片段，可以重建他們?cè)诂F(xiàn)實(shí)生活中的樣子。

The AI's results aren't perfect, but they're pretty good - a remarkable and somewhat terrifying example of how a sophisticated AI can make incredible inferences from tiny snippets of data.
人工智能的結(jié)果并不完美，但它們已經(jīng)相當(dāng)不錯(cuò)了?- 這是一個(gè)細(xì)思恐極例子，說(shuō)明復(fù)雜的人工智能如何從微小的數(shù)據(jù)片段中做出令人難以置信的推斷。

In a paper published this week to the preprint server arXiv, the team describes how it used trained a generative adversarial network to analyze short voice clips and "match several biometric characteristics of the speaker," resulting in "matching accuracies that are much better than chance."
在本周發(fā)布給預(yù)打印服務(wù)器arXiv的一篇論文中，該團(tuán)隊(duì)描述了如何使用經(jīng)過(guò)訓(xùn)練的生成對(duì)抗網(wǎng)絡(luò)來(lái)分析短語(yǔ)音片段并“匹配說(shuō)話者的幾種生物特征”，從而使“匹配準(zhǔn)確性大大提高”。

圖片來(lái)源：視覺(jué)中國(guó)

That's the carefully-couched language of the researchers. In practice, the Speech2Face algorithm seems to have an uncanny knack for spitting out rough likenesses of people based on nothing but their speaking voices.
這是由研究人員精心打造的語(yǔ)言系統(tǒng)。在實(shí)際操作中，Speech2Face算法似乎有一個(gè)神秘的技巧，它只能根據(jù)他們的說(shuō)話聲音產(chǎn)生人們大概的肖像。

The MIT researchers urge caution on the project's GitHub page, acknowledging that the tech raises worrisome questions about privacy and discrimination.
麻省理工學(xué)院的研究人員敦促對(duì)該項(xiàng)目的GitHub頁(yè)面提出警告，承認(rèn)該技術(shù)引發(fā)了關(guān)于隱私和歧視的問(wèn)題令人擔(dān)憂。

"Although this is a purely academic investigation, we feel that it is important to explicitly discuss in the paper a set of ethical considerations due to the potential sensitivity of facial information," they wrote, suggesting that "any further investigation or practical use of this technology will be carefully tested to ensure that the training data is representative of the intended user population."
“雖然這是純粹的學(xué)術(shù)調(diào)查，但我們認(rèn)為，由于面部信息的潛在敏感性，在文章中明確討論一系列的道德因素很重要，”他們寫道，“這表明‘對(duì)此進(jìn)行任何進(jìn)一步調(diào)查或?qū)嶋H應(yīng)用我們都會(huì)對(duì)此進(jìn)行嚴(yán)謹(jǐn)?shù)募夹g(shù)測(cè)試，以確保實(shí)際數(shù)據(jù)能夠代表預(yù)期的用戶群?！?/div>

?

翻譯：進(jìn)擊的Meredith