The Community for Technology Leaders
2013 12th International Conference on Document Analysis and Recognition (2005)
Seoul, Korea
Aug. 31, 2005 to Sept. 1, 2005
ISSN: 1520-5263
ISBN: 0-7695-2420-6
pp: 856-861
Abdul Rahim AHMAD , Laboratoire IRCCyN UMR CNRS Polytech Nantes, France
Christian VIARD-GAUDIN , Laboratoire IRCCyN UMR CNRS Polytech Nantes, France
Emilie CAILLAULT , Laboratoire IRCCyN UMR CNRS Polytech Nantes, France
ABSTRACT
This article analyses the behavior of various hybrid architectures based on a multi-state neuro-markovian scheme (MS-TDNN HMM) applied to online handwriting word recognition systems. We have considered different cost functions, including maximal mutual information criteria with discriminant training and maximum likelihood estimation, to train the systems globally at the word level and also we varied the number of states from one up to three model basic hidden markov models at the letter levekl. We report experimental results for non constrained, write independent, word recognition obtained on the IRONOFF database.
INDEX TERMS
null
CITATION
Abdul Rahim AHMAD, Christian VIARD-GAUDIN, Emilie CAILLAULT, "MS-TDNN with Global Discriminant Trainings", 2013 12th International Conference on Document Analysis and Recognition, vol. 00, no. , pp. 856-861, 2005, doi:10.1109/ICDAR.2005.163
84 ms
(Ver 3.3 (11022016))