Novel Audio Feature Projection Using KDLPCCA-Based Correlation with EEG Features for Favorite Music Classification.
Ryosuke SawataTakahiro OgawaMiki HaseyamaPublished in: IEEE Trans. Affect. Comput. (2019)
Keyphrases
- feature set
- feature vectors
- audio features
- classification accuracy
- feature values
- eeg signals
- feature weights
- feature space
- feature extraction
- feature analysis
- discriminative features
- genre classification
- image features
- feature representation
- individual features
- input features
- feature selection
- feature subset
- irrelevant features
- signal processing
- classification models
- cepstral features
- audio signal
- svm classification
- selecting features
- redundant features
- classification process
- classification method
- spectral features
- textural features
- single feature
- binary features
- svm classifier
- image classification
- highly discriminative
- music information retrieval
- class labels
- multiple features
- extracted features
- music genre classification
- support vector machine svm
- feature ranking
- low level
- weak classifiers
- music score
- feature maps
- speech music discrimination
- audio signals
- machine learning
- invariant features
- pattern recognition
- multimedia
- acoustic features
- physiological signals
- feature selection algorithms
- music collections
- high dimensionality
- decision trees
- automatic music genre classification