Feature Extraction and Database Construction of Vocal Choral Audio Storage on the Basis of Independent Component Analysis

Music art is an important component of human art, and its forms of expression are mainly achieved through various forms such as vocal, choir, and instrumental music. Vocal choir, and audio are important components, but current research and application mainly focus on digital storage based on recording, and there is a lack of research on feature extraction and database construction for audio storage. Based on independent component analysis, this article took vocal choir audio as the research object to address these issues. Through research on feature extraction, library construction, and database construction, a practical and feasible database construction plan was proposed. By analyzing the experimental results, it can be concluded that the highest signal-to-noise ratio of the feature extraction method based on independent component analysis reached 44.8dB. Therefore, feature extraction and database construction based on independent component analysis play an important role in vocal choir audio libraries.