How iir0han/music-genre-classification works
This is a notebook-based music-genre classification project, with notebooks for audio feature extraction, model experiments, and image-based or sequence-model approaches. The tree also contains audio samples, extracted features, and serialized models. The README provides no workflow details, and no source excerpts are available, so relationships between these components cannot be verified and are omitted.
Components 22
Audio and features
- Audio preprocessing
DL for Music genre classification/Preprocessing audio data for Deep Learning - learning how to get the fft,stft,mfcc.ipynb - Segment MFCC extraction
DL for Music genre classification/Preparing the dataset- Extracting mfcc of the segments.ipynb - Feature extraction
features/featuresextraction.ipynb - Final feature extraction
features/Features_Extraction_Final.ipynb - Feature modeling
codes/featureModelling.ipynb - Feature analysis
codes/ImportantFeature.ipynb - Important features
features/ImportantFeatures.ipynb - Librosa tutorial
features/Librosa_tutorial.ipynb
Model experiments
- CNN experiments
DL for Music genre classification/CNN.ipynb - RNN-LSTM experiments
DL for Music genre classification/RNN-LSTM.ipynb - Model building
codes/ModelBuilding.ipynb - Model implementations
codes/ImplementingModels.ipynb - KNN experiment
codes/KNN_Test.ipynb - SVM experiment
codes/SVM_Test.ipynb - Spectrogram images
codes/SpectrogramImages.ipynb
Data and artifacts
- Spectrogram image data
codes/spectogram_image - Audio samples
codes/wav_files - Extracted feature data
features/features_final.csv - Logistic regression model
features/finalized_model_lr.pkl - Random forest model
features/finalized_model_rf.pkl - SVM model
features/finalized_model_svm.pkl - XGBoost model
features/finalized_model_xbg.pkl