Design of an AI Unattended System Based on Home Environment Privacy Protection

This paper focuses on the contradiction between privacy protection and intelligent security in the home environment and proposes a solution based on an AI unattended system. By integrating pose estimation and an improved ST-GCN algorithm, combined with edge AI processing technology, an architecture combining infrared imaging hardware and lightweight deep neural networks is constructed to achieve data edge autonomy and reduce the risk of privacy leakage. The system adopts multiple privacy protection measures in data acquisition, processing, and storage. Experimental verification shows that the behavior recognition accuracy reaches 97.2%, meeting the real-time and stability requirements of large-scale monitoring, and providing a safe and reliable protection solution for families.