阿里巴巴集团近日推出了一项名为Mobile-Agent的手机操作智能体框架技术。据最新发布的论文显示，这一框架具备了操作多达10款应用软件的能力，并能在不同应用间无缝切换，完成用户所赋予的任务。令人瞩目的亮点是，Mobile-Agent框架无需事先训练，便能即插即用，而且整个操作过程基于视觉能力实现，不再需要为每个应用编写繁琐的XML操作文档。
举例来说，当用户给出指令，Mobile-Agent可以自主搜索篮球比赛的结果，并根据比赛的具体情况，在备忘录中自动撰写相关的文稿。这一创新技术打破了传统应用之间的界限，实现了跨应用的任务执行，使得手机真正成为了功能强大的智能助手。
The Alibaba Group recently launched a new mobile operating intelligent agent framework technology named Mobile-Agent. According to the latest published paper, this framework has the ability to operate up to 10 different application software and switch seamlessly between them to complete tasks assigned by users. A remarkable highlight is that the Mobile-Agent framework does not require prior training and can be used immediately upon insertion. The entire operation process is based on visual capabilities, eliminating the need for writing XML operation documents for each application.
For example, when users give an instruction, Mobile-Agent can autonomously search for basketball game results and automatically draft related documents in the memo based on the game’s situation. This innovative technology breaks the boundaries between traditional applications, enabling cross-application task execution and turning mobile phones into powerful intelligent assistants.
【来源】https://www.qbitai.com/2024/02/118426.html
Views: 1