Abstract: With the popularization of artificial intelligence, especially the rapid application and promotion of AIGC technology, it is bound to have a significant impact on intelligent interactive experiences. For instance, the original smart home scenes were preset and configured by various manufacturers, and users could only passively use and accept fixed combinations of smart home scenes. Moreover, most users are unfamiliar with the configuration methods of smart appliances, which fails to meet their needs for personalized smart scene customization. Focusing on the intelligent home field, and considering the various experiential problems users encounter while using whole house smart scenes, this method provides a solution that combines AIGC large model technology to achieve self-arrangement of whole house smart scenes. This allows users to easily customize smart scenes by either speaking a command or dragging devices on a screen, making it convenient and fast. This effectively addresses the long-standing issues of users not knowing how to use the systems and stores being reluctant to promote them. The ultimate goal is to create a new engine for whole house smart scene interaction, so that users' smart scene experiences are unique and innovative, frequently updated and tailored to individual needs.

Key words: AIGC large-scale model, Smart scenario self-orchestration, Intelligent interaction engine