Keyphrases
- similarity measure
- error rate
- main contribution
- probabilistic model
- high precision
- preprocessing
- significant improvement
- cost function
- detection method
- computational complexity
- dynamic programming
- support vector machine svm
- data sets
- segmentation method
- synthetic data
- optimization method
- fully automatic
- clustering method
- theoretical analysis
- computationally efficient
- em algorithm
- computational cost
- prior knowledge
- k means
- objective function
- multiscale
- computer vision
- information retrieval