A Real-Time Hand-Gesture Recognition Using Deep Learning Techniques

Hand sign recognition is an important technology in human-computer interaction because it allows people to communicate with machines in a natural and simple manner. This paper offers a revolutionary real-time hand sign identification method using MediaPipe landmark methods and OpenCV. This system uses a deep learning and computer vision combination to accurately understand and classify a large range of hand gestures and signs. This technique can be applied in many different contexts, particularly in the domains of immersive technologies (Virtual and Augmented Reality), and sign language interpretation. This project is based on MediaPipe, a library that enables for the real-time monitoring of hand landmarks in video streams. These landmarks stand in for important hand locations including the palm, knuckles, and fingertips. From the landmark data, significant features are extracted and used as input to a machine learning model to perform the classification task. The outcomes validate the system’s precision and show that it can operate in real-time, which qualifies it for interactive applications where hand sign recognition is crucial. There are plenty of applications using hand gesture recognition like photo snapping, video games and application control systems. This method makes an important contribution to the fields of computer vision and human-computer interaction by providing a realistic and efficient solution for hand sign recognition. Ten gestures have been considered and we were able to achieve promising results for almost all of the gestures.