Optimized leaky ReLU for handwritten Arabic character recognition using convolution neural networks.
Bahera H. NayefSiti Norul Huda Sheikh AbdullahRossilawati SulaimanZaid Abdi Alkareem AlyasseriPublished in: Multim. Tools Appl. (2022)
Keyphrases
- neural network
- handwritten words
- writer independent
- character recognition
- writer identification
- handwritten characters
- optical character recognition
- handwriting recognition
- character segmentation
- handwritten word recognition
- handwritten documents
- recognition rate
- pattern recognition
- printed text
- printed characters
- printed documents
- word recognition
- ocr systems
- hidden markov models
- arabic documents
- training process
- genetic algorithm
- neural nets
- recurrent neural networks
- word spotting
- hand written
- neural network model
- writing style
- writing styles
- artificial neural networks
- back propagation
- document images
- license plate
- machine vision
- multilayer perceptron
- document image analysis
- multi layer
- fuzzy systems
- automatic recognition
- feed forward
- handwritten text
- digital libraries
- arabic text