A Study on the Classification and Inheritance Genealogy of Ethnic Music Intangible Cultural Heritage Based on Audio Signal Processing

Accurate style identification, lineage reconstruction, and digital preservation remain critical challenges in safeguarding ethnic music intangible cultural heritage. As audio signal processing and intelligent sensing technologies continue to evolve, efficient feature extraction and pattern recognition methods have become increasingly important for multimedia information analysis and electromagnetic signal processing applications. This study proposes a framework for ethnic music style classification and inheritance genealogy reconstruction based on audio signal processing. Mel-Frequency Cepstral Coefficients (MFCCs) are first extracted to characterize acoustic features, after which a hybrid Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM) architecture is employed for style classification. Subsequently, Dynamic Time Warping (DTW) is adopted to measure the similarity between musical segments, and hierarchical clustering is utilized to reconstruct the lineage network. A comprehensive evaluation dataset containing 2,847 audio samples from 12 ethnic categories is established for validation. Experimental results demonstrate that the proposed approach achieves a style classification accuracy of 94.3% and successfully identifies 89 inheritance branches, accurately reconstructing the evolutionary relationships of ethnic music traditions, including Tibetan and Mongolian genres. The proposed framework provides an effective technical solution for the digital protection and inheritance of ethnic music intangible cultural heritage while offering methodological reference for intelligent acoustic sensing, electromagnetic signal analysis, and multimodal information processing systems.