Problem Statement :Our project is focused on developing a web-based hand gesture recognition system for video control applications. It allows users to interact with video content through intuitive real time gestures for playback actions like play, pause, fast-forward, all via a browser extension. Its goal is to enhance user interaction, providing a hands-free and accessible experience while ensuring accuracy and responsiveness
Objectives : Provide an accessible, hands-free interface for users, improving the multimedia experience by creating a web-based hand gesture recognition system to control video applications using simple hand movements.
Methodology : Developed a gesture recognition system using deep learning and computer vision. It utilizes OpenCV for video capture and preprocessing and MediaPipe for hand landmark detection. A Convolutional Neural Network (CNN) was trained on a diverse dataset of hand gesture images, incorporating data augmentation techniques to improve generalization. The system is implemented using a Flask-based API for real-time gesture classification, which communicates with a Google Chrome extension to execute video playback commands.