Vocal music teaching evaluation model based on pattern recognition and voiceprint feature analysis

Vocal practice in vocal music teaching is a purposeful and planned perceptual process, which has reached the goal of becoming proficient and forming the motive force of singing. Proficiency is formed after repeated conscious actions. There are many factors that affect the performance of a piece of music, and there are also many evaluation indexes, such as rhythm, expressive force, music sense and style. Neural network is a mathematical model proposed by simulating the thinking mode of human brain in artificial intelligence. It has the advantages of not strict data distribution requirements, nonlinear data processing methods, strong robustness and dynamics, and is very suitable as a mathematical model of evaluation system. This paper proposes a vocal music teaching evaluation model based on pattern recognition voiceprint feature analysis, which extracts the voiceprint features of the purified music speech signal and recognizes music according to the voiceprint feature extraction results. Simulation results show that the accuracy of speech feature extraction using this method is good, and the resolution of music recognition is high.