Application Research of Neural Network Technology in Vocal Music Evaluation

In order to solve the problem that vocal performers are greatly influenced by subjective factors in scoring, this paper takes vocal music evaluation system as input and adopts BP neural network to establish evaluation model. We provide the basic structure and application process of BP neural network in residual evaluation, and propose the method of adaptive learning factor and elastic gradient and adding momentum term to improve BP algorithm. After preprocessing, the extracted feature parameters are used as the input variables of neural network for the next step of training. Finally, MATLAB is used to analyze and process the collected data, and the results are compared with the subjective evaluation method. The simulation results show that the proposed scheme significantly improves the accuracy and robustness of music classification and recognition, and the system can reflect the real level of performers.