Abstract
The interaction between man and computer is expanded with computer technology development. Hand gesture recognition is an essential technology for human-computer interaction, and it allows the recording and interpretation of hand gestures without some devices' use. Since the hand is a man's primary working tool, its position and orientation in space are crucial for many potential applications, such as interaction with virtual reality devices. Hand position estimation is a substantial process of hand gesture recognition. However, there are many difficulties in 3D hand pose estimating from a single RGB image. Media Pipe is a high-quality finger and hand-tracking framework. It uses in machine learning to detect 3D landmarks. The proposed paper provides a real-time approach to recognizing Kazakh gestures performance using a conventional camera and Media Pipe.