A MRF model with parameter optimization by CRF for on-line recognition of handwritten Japanese characters.
Bilan ZhuMasaki NakagawaPublished in: DRR (2011)
Keyphrases
- parameter optimization
- markov random field
- mrf model
- handwritten characters
- parameter estimation
- character recognition
- character segmentation
- conditional random fields
- handwritten words
- handwritten text
- random fields
- chinese characters
- handwriting recognition
- maximum a posteriori
- genetic algorithm ga
- pairwise
- differential evolution
- license plate
- word recognition
- hyperparameters
- graph cuts
- higher order
- genetic algorithm
- image segmentation
- energy function
- writer independent
- bayesian framework
- text lines
- hidden markov models
- parameter settings
- feature extraction
- support vector machine
- computer vision
- parameter values
- active learning
- similarity measure