CNPM: Chinese National Pentatonic Mode Recognition Model
| Property | Value |
|---|---|
| Author | ccmusic-database |
| Best Performance | 85.9% (VIT-L-16 with CQT) |
| Model Size Range | 7.4M - 306.5M parameters |
| Dataset | Chinese National Pentatonic Mode Dataset |
What is CNPM?
CNPM is a sophisticated machine learning model designed specifically for recognizing and classifying traditional Chinese pentatonic modes. The model can identify five primary tonal modes in Chinese music: Gong, Shang, Jiao, Zhi, and Yu, covering five-tone, six-tone, and seven-tone scales. It represents a significant advancement in the digital preservation and analysis of ethnic music.
Implementation Details
The model implements various architectures including ViT, VGG, RegNet, and ResNet variants, with performance evaluated across different input representations (Mel, CQT, and Chroma). The best performing configuration uses a ViT-L-16 architecture achieving 85.9% accuracy with CQT input features.
- Advanced feature extraction and spectral analysis capabilities
- Multiple backbone architecture options for different performance needs
- Comprehensive evaluation across different input representations
- Efficient processing of traditional Chinese music audio files
Core Capabilities
- Accurate recognition of five traditional Chinese pentatonic modes
- Support for analysis of five-tone, six-tone, and seven-tone scales
- Digital preservation of ethnic music characteristics
- Robust pattern recognition for musical mode classification
Frequently Asked Questions
Q: What makes this model unique?
CNPM is specifically designed for Chinese pentatonic music analysis, combining traditional musicology with modern AI techniques. It offers state-of-the-art accuracy in recognizing traditional Chinese modes, making it invaluable for music preservation and analysis.
Q: What are the recommended use cases?
The model is ideal for musicologists studying Chinese traditional music, digital archives working on music preservation, and researchers analyzing ethnic music patterns. It can be used for automated classification of music pieces, educational purposes, and cultural heritage preservation.
