Treffer: GECO: A Real-Time Computer Vision-Assisted Gesture Controller for Advanced IoT Home System.
Sensors (Basel). 2024 Nov 10;24(22):. (PMID: 39598974)
IoTDI 2017 (2017). 2017 Apr;2017:47-58. (PMID: 29683151)
Sensors (Basel). 2024 Apr 10;24(8):. (PMID: 38676024)
Sci Rep. 2025 May 1;15(1):15284. (PMID: 40312474)
PLoS One. 2025 Apr 24;20(4):e0319996. (PMID: 40273209)
Sensors (Basel). 2025 Jan 06;25(1):. (PMID: 39797073)
Nat Methods. 2020 Mar;17(3):261-272. (PMID: 32015543)
Weitere Informationen
This paper introduces GECO, a real-time, computer vision-assisted gesture controller for IoT-based smart home systems. The platform uses a markerless MediaPipe interface that combines gesture-driven navigation and command execution, enabling intuitive control of multiple domestic devices. The system integrates binary and analog gestures, such as continuous light dimming based on thumb-index angles, while operating on-device through a private MQTT network. Technical evaluations across multiple Android devices have demonstrated ultra-low latency times (<50 ms), enabling real-time responsiveness. A user experience study with seventeen participants reported high intuitiveness (9.5/10), gesture accuracy (9.2/10), and perceived inclusivity, mainly for individuals with speech impairments and low technological literacy. These results position GECO as a lightweight, accessible, and privacy-preserving interaction framework, advancing the integration of artificial intelligence and IoT within smart home environments.