A Method of Combining Gaussian Mixture Model and K-Means for Automatic Audio Segmentation of Popular Music.
Ing-Jr DingPublished in: ICITCS (2012)
Keyphrases
- gaussian mixture model
- segmentation method
- k means
- em algorithm
- segmentation algorithm
- gaussian distribution
- expectation maximization
- clustering method
- similarity measure
- pairwise
- unsupervised learning
- detection method
- learning algorithm
- gaussian model
- background subtraction
- covariance matrix
- statistical model
- feature space
- gaussian mixture
- information retrieval
- speaker recognition