Error-Correcting Output Coding for the Convolutional Neural Network for Optical Character Recognition.
Huiqun DengGeorge StathopoulosChing Y. SuenPublished in: ICDAR (2009)
Keyphrases
- optical character recognition
- convolutional neural network
- error correcting output coding
- web page prediction
- multi class
- multiclass classification
- face detection
- character recognition
- ocr systems
- error correcting output codes
- document images
- prediction accuracy
- text recognition
- multi class classification
- conditional random fields
- neural network
- multi class learning
- text classification
- binary classification
- multiclass problems
- scanned documents
- printed documents
- text extraction
- prefetching
- binary classifiers
- cost sensitive
- hidden markov models
- feature space
- bayesian networks
- handwriting recognition
- machine vision
- response time
- higher order
- object detection
- pairwise
- historical manuscripts
- learning algorithm
- machine learning