The integration of artificial intelligence and ethnic music cultural inheritance under deep learning

The traditional music education system faces numerous challenges in inheriting ethnic music culture. Especially in the modern educational environment, the protection and dissemination of ethnic music encounter many difficulties. This work aims to utilize advanced technologies such as deep learning (DL) to explore methods for optimizing the inheritance of ethnic music culture. By summarizing the current situation of ethnic music cultural inheritance, and analyzing its background and content, this work proposes innovative solutions that integrate artificial intelligence (AI) technology. Leveraging a newly constructed model, it performs multi-level and comprehensive analyses of ethnic music elements, uncovering the internal emotional expression mechanism of ethnic music. The experimental results of timbre emotion recognition are presented and compared. The findings reveal that the unsupervised training method improves the feature accuracy by 1.96% compared to the Mel-Frequency Cepstral Coefficients (MFCC), while the supervised training method achieves a 3.46% improvement. In addition, the timbre recognition rate is compared between the Gaussian Mixture Model-Hidden Markov Model (GMM-HMM) and the Deep Neural Network-Hidden Markov Model (DNN-HMM). The result shows that the DNN-HMM is better. These findings highlight the significant advantages of applying DL methods in preserving and transmissing ethnic music cultural inheritance. This work can effectively enhance the accuracy of music emotion recognition, thus providing new technical support for the protection and inheritance of ethnic music.