USING CONVOLUTIONAL NEURAL NETWORKS FOR POSE ESTIMATION AND GESTURE RECOGNITION

The central theme of this research is the application of advanced artificial intelligence technologies for analyzing human hand movements using data obtained from optical sensors such as cameras. The main goal is the development of effective algorithms and machine learning models capable of identifying, tracking, and analyzing the hand’s skeletal structure, with subsequent gesture recognition for intuitive remote control of electronic devices.

The aim of this work is a deep exploration and systematic comparison of existing methodologies and technological approaches to solving the complex task of gesture recognition. This includes analysis of individual modules that play a critical role in the gesture recognition process, in particular the pose estimation modules which allow precise interpretation of user movements, and specialized recognition algorithms that can accurately identify specific gestures based on the collected data.