Off-Line Handwritten Arabic Word Recognition Using SVMs with Normalized Poly Kernel.
Abdulrahman AlalshekmubarakAmir HussainQiu-Feng WangPublished in: ICONIP (2) (2012)
Keyphrases
- handwritten words
- support vector
- handwritten documents
- arabic documents
- kernel function
- sparse kernel
- word recognition
- word spotting
- kernel methods
- handwriting recognition
- writer independent
- kernel space
- numeral strings
- character recognition
- feature space
- printed text
- hidden markov models
- rbf kernel
- support vector machine
- polynomial kernels
- printed documents
- kernel parameters
- word segmentation
- kernel machines
- support vectors
- kernel principal component analysis
- multiple kernel learning
- feature selection
- linear svm
- character n grams
- gaussian kernel
- input space
- document analysis
- kernel matrix
- recognition rate
- svm classifier
- radial basis function
- hyperplane
- kernel pca
- optical character recognition
- maximum margin
- decision function
- similarity measure
- document images
- n gram
- unknown words
- machine learning
- text lines
- writing style
- co occurrence
- speech recognition
- document level
- writer identification
- large margin classifiers
- machine vision
- tree kernels
- standard svm
- gaussian kernels