近日，谷歌研究人员推出了一款名为Genie的全新AI模型，该模型能够通过一张图片生成一个交互式的虚拟世界。Genie拥有110亿参数，它使生成的世界具有“动作可控”的特点，用户可以在其中进行逐帧行动。谷歌将Genie定义为“生成式AI的一种新模式”，并认为它将开启“图/文生成交互世界”的新时代，同时成为实现通用AI Agent的催化剂。Genie的问世标志着AI技术在交互式内容生成领域的又一重大突破，为未来的虚拟现实、游戏设计和教育培训等领域提供了新的可能性。
英文标题：Google Genie: One Image Generates Interactive World
英文关键词：Google AI, Interactive World, General AI Agent
英文新闻内容：
Google researchers have recently unveiled a new AI model called Genie. With 110 billion parameters, Genie can generate an interactive virtual world from a single image. The model allows users to navigate and control actions within the generated world frame by frame. Google has defined Genie as a “new paradigm of generative AI” and believes it will usher in a new era of “image/text-based interactive worlds.” The tech giant also sees Genie as a catalyst for the realization of a general AI agent. The introduction of Genie signifies a significant breakthrough in AI technology for interactive content generation, offering new possibilities for virtual reality, game design, and educational training.
【来源】https://www.chinastarmarket.cn/detail/1604953
Views: 1