An Approach Of Hand Gestures Recognition For Household Appliances Administration Using Convolutional Neural Network

Abstract

In the swiftly evolving landscape of machine learning applications, their integration into various facets of human life, notably within household appliances, has witnessed notable growth. Accompanying this proliferation requires practical control tools and methods, which often encounter challenges related to complexity and limited commercial applicability despite the convenience they promise. This thesis endeavors to address this predicament by presenting a novel approach to hand gesture recognition for the administration of household appliances. The methodology aims to overcome the shortcomings of current techniques by offering a lightweight, user-friendly, and expeditious solution. The proposed system leverages state-ofthe-art technologies such as MediaPipe, Tensorflow, OpenCV, and a Convolutional Neural Network architecture. By amalgamating these tools, the research seeks to provide an accessible and efficient means of controlling appliances through intuitive hand gestures, thereby enhancing user experience, and circumventing the complexities and commercial limitations inherent in current methods.