Performances of low-level audio classifiers for large-scale music similarity.
Julien OsmalskyjMarc Van DroogenbroeckJean-Jacques EmbrechtsPublished in: IWSSIP (2014)
Keyphrases
- audio features
- low level
- music recommendation
- feature set
- polyphonic music
- audio signals
- high level
- automatic music genre classification
- music information retrieval
- audio visual
- audio signal
- music score
- genre classification
- music retrieval
- visual information
- music scores
- visual features
- music genre classification
- similarity measure
- classifier training
- higher level
- support vector
- audio recordings
- training data
- acoustic features
- feature selection
- audio content
- feature vectors
- music collections
- training set
- speaker identification
- decision trees
- low level image features
- audio files
- multimedia
- multi modal
- sound source
- visual cues
- naive bayes
- classification performances
- similarity estimation
- class labels
- classification algorithm
- low level features
- digital audio
- extracted features
- distance measure
- feature space
- digital music
- hidden markov models
- multiple classifiers
- speech music discrimination
- machine learning