95 articles 

inria-00112630, version 1

On the Accent in Handwriting of Individuals

Faisal Farooq () 1, Liana Lorigo () 1, Venu Govindaraju () 1

Tenth International Workshop on Frontiers in Handwriting Recognition (2006)

Abstract: This paper introduces the novel concept of accent in handwritten text. Accent in speech has been studied by linguists as well as computer scientists, however, it has not been considered in handwriting. We validate this concept in handwriting. We perform our experiments with two sets of Arabic writers. The Native writers are whose first written script is Arabic and Non-Natives that wrote other scripts before learning Arabic as an additional script. Outputs from directional Gabor filters were used as input features to an SVM for classification. We achieved an overall accuracy of 86.96% and 78.54% in two different settings at classifying a given word as being from either class. A perfect classification at a document level was achieved for our dataset using a max-vote scheme.

  • 1:  Center of Excellence for Document Analysis and Recognition (CEDAR)
  • State University of New York at Buffalo
  • Domain : Computer Science/Document and Text Processing
    Computer Science/Computer Vision and Pattern Recognition
  • Keywords : Accent – Handwriting – Gabor filters – SVM
  • Comment : http://www.suvisoft.com
 
  • inria-00112630, version 1
  • oai:hal.inria.fr:inria-00112630
  • From: 
  • Submitted on: Thursday, 9 November 2006 13:41:47
  • Updated on: Monday, 19 March 2007 09:41:19