Hierarchical hybrid MLP/HMM or rather MLP features for a discriminatively trained Gaussian HMM: A comparison for offline handwriting recognition
- 1 September 2011
- conference paper
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 3541-3544
- https://doi.org/10.1109/icip.2011.6116480
Abstract
We use neural network based features extracted by a hierarchical multilayer-perceptron (MLP) network either in a hybrid MLP/HMM approach or to discriminatively retrain a Gaussian hidden Markov model (GHMM) system in a tandem approach. MLP networks have been successfully used to model long-term and non-linear features dependencies in automatic speech and optical character recognition. In offline hand writing recognition, MLPs have been mostly used for isolated character and word recognition in hybrid approaches. Here we analyze MLPs within an LVCSR framework for continuous handwriting recognition using discriminative MMI/MPE training. Especially hybrid MLP/HMM and discriminatively retrained MLP-GHMM tandem approaches are evaluated. Significant improvements and competitive results are re ported for a closed-vocabulary task on the IfN/ENIT Arabic handwriting database and for a large-vocabulary task using the IAM English handwriting database.Keywords
This publication has 13 references indexed in Scilit:
- Confidence- and margin-based MMI/MPE discriminative training for off-line handwriting recognitionInternational Journal on Document Analysis and Recognition (IJDAR), 2011
- ICFHR 2010 - Arabic Handwriting Recognition CompetitionPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2010
- Margin-Based Discriminative Training for String RecognitionIEEE Journal of Selected Topics in Signal Processing, 2010
- Improving Offline Handwritten Text Recognition with Hybrid HMM/ANN ModelsIeee Transactions On Pattern Analysis and Machine Intelligence, 2010
- Modified MPE/MMI in a transducer-based frameworkPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2009
- ICDAR 2009 Arabic Handwriting Recognition CompetitionPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2009
- Hidden Markov model-based ensemble methods for offline handwritten text line recognitionPattern Recognition, 2008
- Minimum classification error training of hidden Markov models for handwriting recognitionPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2002
- The IAM-database: an English sentence database for offline handwriting recognitionInternational Journal on Document Analysis and Recognition (IJDAR), 2002
- Connectionist Speech RecognitionPublished by Springer Science and Business Media LLC ,1994