Virtual Mouse Using Hand Gesture And Voice Recognition

Abstract

One of the most widely used tools for human-computer interaction is the mouse, which comes in three main varieties: wired, wireless, and Bluetooth. All of these, meanwhile, necessitate a physical connection—typically through a dongle— which makes device setup more difficult. We suggest a brand-new hand gesture- based cursor control method that does away with conventional hardware. Rather, it uses computer vision and machine learning. Convolutional neural networks (CNNs) implemented with MediaPipe are used in the suggested system to provide a reliable, hardware-independent gesture recognition technique. This technique expands on S. Shriram's approach by allowing users to manipulate clicks, drag objects, and move the pointer using a variety of hand gestures. Python and OpenCV provide the foundation of the system, which simply needs a computer and a camera.