VOA科学技术2023 谷歌发布数字水印工具识别人工智能生成图像

（单词翻译）

Google Launches Watermark Tool to Identify AI-created Images

谷歌发布数字水印工具识别人工智能生成图像

Google has launched a tool designed to mark identity on images created by artificial intelligence (AI) technology.

谷歌推出了一款工具，旨在给人工智能生成的图像标记身份。

The American-based search engine and online advertising¹ company announced the new tool in a statement Tuesday. Google has already made the system available to a limited number of beta testers. Google plans to expand its use in the future.

这家总部位于美国的搜索引擎和网络广告公司周二在一份声明中宣布了这一新工具。谷歌已经向数量有限的测试人员提供了该系统。谷歌计划在未来扩大其使用范围。

A team at Google Deep Mind developed the tool, called SynthID, in partnership² with Google Research. SynthID is designed to work with Google's AI image creator Imagen. That system creates full film-quality images from simple text commands. For now, only Imagen users will be able to use the AI identifying tool.

谷歌Deep Mind的一个团队与谷歌研究院合作开发了这款名为SynthID的工具。SynthID旨在与谷歌旗下的人工智能图像生成器Imagen搭配使用。该系统可以通过简单的文本命令生成全胶片质量的图像。目前，只有Imagen用户才能使用这款人工智能识别工具。

SynthID works by creating a hidden, digital "watermark" on images. Watermarks have long been used with paper documents and money as a way to mark them as being real, or authentic³. With this method, paper can be held up to a light to see if a watermark exists and the document is authentic.

SynthID的工作原理是在图像上创建一个隐藏的数字水印。水印长期以来一直被用于纸质文件和货币，作为将其标记为真实的一种方式。使用这种方法，可以将纸张举到灯光下，查看水印是否存在以及文档是否真实。

A digital watermark serves the same purpose. But it uses technology to digitally embed⁴ hidden markers in images. These watermarks are hard to see by just looking at an image. Instead, special tools are needed to identify the watermark.

数字水印也有同样的作用。但它是使用技术在图像中数字化嵌入隐藏的标记。肉眼很难看到这些水印。相反，识别这种水印需要特殊的工具。

Google says the digital watermark is designed to help individuals and companies identify whether an image has been created by AI tools or not. This could help people recognize inauthentic pictures published online and also protect copyright-protected images.

谷歌表示，数字水印旨在帮助个人和公司识别图像是否由人工智能工具生成。这可以帮助人们识别网上发布的不真实图片，也可以保护受版权保护的图片。

Technology experts have identified these issues as two of the biggest problems with AI creation tools – they can increase the amount of misinformation online and they can violate copyrights.

技术专家认为这是人工智能创作工具的两个最大问题——它们可能会增加网上错误信息的数量，也可能侵犯版权。

Last month, Google's parent Alphabet joined other major technology companies in agreeing to establish watermark tools to help make AI technology safer.

上个月，谷歌母公司Alphabet与其他大型科技公司一道，同意创建水印工具，以帮助提高人工智能技术的安全性。

The companies – including Facebook parent Meta, Microsoft-backed AI developer OpenAI and online seller Amazon – made the agreement during a meeting with U.S. President Joe Biden at the White House. They promised to fully⁵ develop and test watermarking tools for content including text, audio and video.

包括脸书母公司Meta、微软支持的人工智能开发商OpenAI以及网上卖家亚马逊在内的这些公司在白宫与美国总统乔·拜登会面时达成了协议。他们承诺将全面开发和测试文本、音频和视频等内容的水印工具。

Google said its system embeds⁶ a digital watermark directly into the details of an image. The tool can be used for "connecting people with high-quality information and upholding trust between creators and users across society."

谷歌表示，其系统将数字水印直接嵌入图像的细节中。该工具可用于“将人们与高质量的信息联系起来，维护全社会创作者和用户之间的信任”。

Developers of the SynthID system said it is built to keep the watermark in place even if the image itself is changed by creative tools designed to resize pictures or add additional light or color.

SynthID系统的开发人员表示，即使图像本身被设计用于调整图片大小或添加额外光线或颜色的创造性工具更改，它也能保持水印不消失。

The tool can add a hidden watermark to AI-produced images created by Imagen. SynthID can also examine an image to find a digital watermark that was embedded⁷ with the Imagen system.

该工具可以为Imagen生成的人工智能图像添加隐藏水印。SynthID还可以检查图像，找出Imagen系统所嵌入的数字水印。

Google says it will continue to test the watermarking tool and hopes to collect many user experiences from the current beta testers. And the company looks forward to adding the system to other Google products and making it available to more individuals and organizations.

谷歌表示，它将继续测试这款水印工具，并希望从目前的测试人员那里收集到许多用户体验。该公司期待着将该系统添加到谷歌的其他产品中，并将其提供给更多的个人和组织。

In addition to SynthID, Google also announced Tuesday the launch of additional AI tools designed for businesses and structural⁸ improvements to its computing⁹ systems. Those systems are used to produce AI tools, also known as large language models.

除了SynthID，谷歌周二还宣布推出更多为企业设计的人工智能工具，并对其计算系统进行结构改进。这些系统用于生成也称为大型语言模型的人工智能工具。

Google made public a new version of its self-designed AI computing chips. The company says the new chip, called TPU v5e, was built to train large computer models, but also more effectively serve those models. Google also released new versions of software and security tools designed to work with AI systems.

谷歌公开了其自行设计的新版人工智能计算芯片。该公司表示，这款名为TPU v5e的新芯片旨在训练大型计算机模型，但也能更有效地为这些模型服务。谷歌还发布了新版本的软件和安全工具，旨在与人工智能系统配合使用。