• DocumentCode
    3475300
  • Title

    Analysis of timing constraint on combined SVM-HMM classifier and SVM classifier

  • Author

    Kumawat, Ashish Kumar ; Khandelwal, Sourabh

  • Author_Institution
    Comput. Sci. Eng., Geetanjali Inst. of Tech. Studies, Udaipur, India
  • fYear
    2013
  • fDate
    20-22 Dec. 2013
  • Firstpage
    214
  • Lastpage
    218
  • Abstract
    In Handwriting Verifier Timing constraint is very crucial part which have used in the SVM Classifier, these have using the more time for the large number of sample and get the less accuracy. When there is used the Combined SVM-HMM so that has taken less time for the analysis the large sample and give better accuracy than SVM. That all time has evaluated with the curvelet transform and make a digital clock pulse in form of 1´s and 0´s. Which have calculate in the invariant movement with the wavelength from the trained data image and replace them, on the place of selected writing image, than make an metric of binary number and calculate them with the method of invariant curve let, thereafter compare the character on the time of the calculate image binary code and make an metrics on the striate line of SVM-HMM in terms of 1´s and 0´s.
  • Keywords
    binary codes; curvelet transforms; handwriting recognition; hidden Markov models; image classification; image coding; support vector machines; SVM-HMM classifier; binary number; curvelet transform; digital clock pulse; handwriting verifier timing constraint; image binary code; invariant curvelet; invariant movement; writing image; Conferences; Decision support systems; Technological innovation; US Department of Defense; Digital Clock pulse; Hidden markov modal (HMM); Support vector machine (SVM);
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    MOOC Innovation and Technology in Education (MITE), 2013 IEEE International Conference in
  • Conference_Location
    Jaipur
  • Print_ISBN
    978-1-4799-1625-2
  • Type

    conf

  • DOI
    10.1109/MITE.2013.6756337
  • Filename
    6756337