Research on the construction of virtual playing environment and its human-computer interaction performance improvement in piano teaching

Computer vision based gesture recognition is widely used in the field of human-computer interaction. It is able to use relevant algorithms to analyze the image sequence captured by the camera and recognize the gestures in it. This paper mainly focuses on somatosensory interaction technology and mixed reality technology to construct a virtual playing environment for piano teaching. First, this paper proposes a gesture recognition algorithm based on the Hidden Markov Model (HMM), which uses the Baum-Welch algorithm to train the model and improve the performance of human-computer interaction. Then, the piano is modeled in three dimensions, and the design of the virtual piano playing scheme is realized by combining mixed reality technology. Finally, the application analysis of the model is performed. The recognition rate of the HMM for both gestures of extended two fingers and half clenched fist is greater than 97%, and the recognition rate of gestures by the model is high. In addition, the questionnaire research on the supporting product of piano virtual performance shows that the product enjoys wide popularity among users, and the users evaluate its interaction experience at all levels highly, but there still exists some room for improvement and enhancement, which provides a strong basis for the further improvement of the performance of human-computer interaction in the virtual performance of piano teaching.