Maximum mutual information training for an online neural predictive handwritten word recognition system

Garcia-Salicetti, Sonia; Dorizzi, Bernadette; Gallinari, Patrick; Wimmer, Zsolt

doi:10.1007/pl00013574

Cited by 7 publications

(4 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Although Latin-alphabet based scripts have been getting much of the attention, the number of studies on recognition of many other writing systems increased recently [8], [2], [3], [42]. 2 Different machine learning techniques like Hidden Markov Models (HMM), Support Vector Machines (SVM) and Artificial Neural Networks (ANN) and their combinations are employed in online handwritten text recognition systems in the literature [9], [10], [11], [12], [13], [14], [15], [16]. HMM based recognizers have been particularly very popular due to their capability of modelling time series effectively [17], [18], [19], [20].…”

Section: Related Work (İli̇şki̇li̇ çAlişmalar)mentioning

confidence: 99%

Recognition of Online Turkish Handwriting using Transfer Learning

BİLGİN TAŞDEMİR

2023

Gazi Üniversitesi Fen Bilimleri Dergisi Part C: Tasarım Ve Teknoloji

View full text Add to dashboard Cite

We present a recognition system for online Turkish handwriting using transfer learning. Training deep networks requires large amounts of data. Since such a sufficiently large collection of Turkish handwriting samples is not available, So we adopt the transfer learning approach and train and optimize a CNN-BLSTM recognition system first using the standard IAM-On dataset of English handwriting. Then, we fine tune it with Turkish handwriting samples from a smaller dataset. Fine tuning increases the character recognition rate of the final system which is evaluated on 2,041 samples of isolated Turkish words from the initial value of 49% to 85%. The results show that transfer learning can be a solution to the data scarcity problem of online Turkish handwriting.

show abstract

Section: Related Work (İli̇şki̇li̇ çAlişmalar)mentioning

confidence: 99%

Recognition of Online Turkish Handwriting using Transfer Learning

BİLGİN TAŞDEMİR

2023

Gazi Üniversitesi Fen Bilimleri Dergisi Part C: Tasarım Ve Teknoloji

View full text Add to dashboard Cite

show abstract

“…HMMs are also used in hybrid systems together with different kinds of artificial neural networks (ANNs), including deep learning methods such as recurrent neural networks (RNNs) and LSTMs [15][16][17][18][19][20][21][22][23]. For instance, ANNs are employed for extending the HMM with contextual information [15,16], feature extraction [17,20], and predicting emission probability densities in an HMM system [19,21].…”

Section: Related Workmentioning

confidence: 99%

Large vocabulary recognition for online Turkish handwriting with sublexical units

Taşdemir¹,

Yanıkoğlu²

2018

Turk J Elec Eng & Comp Sci

View full text Add to dashboard Cite

We present a system for large vocabulary recognition of online Turkish handwriting, using hidden Markov models. While using a traditional approach for the recognizer, we have identified and developed solutions for the main problems specific to Turkish handwriting recognition. First, since large amounts of Turkish handwriting samples are not available, the system is trained and optimized using the large UNIPEN dataset of English handwriting, before extending it to Turkish using a small Turkish dataset. The delayed strokes, which pose a significant source of variation in writing order due to the large number of diacritical marks in Turkish, are removed during preprocessing. Finally, as a solution to the high out-of-vocabulary rates encountered when using a fixed size lexicon in general purpose recognition, a lexicon is constructed from sublexical units (stems and endings) learned from a large Turkish corpus. A statistical bigram language model learned from the same corpus is also applied during the decoding process. The system obtains a 91.7% word recognition rate when tested on a small Turkish handwritten word dataset using a medium sized (1950 words) lexicon corresponding to the vocabulary of the test set and 63.8% using a large, general purpose lexicon (130,000 words). However, with the proposed stem+ending lexicon (12,500 words) and bigram language model with lattice expansion, a 67.9% word recognition accuracy is obtained, surpassing the results obtained with the general purpose lexicon while using a much smaller one.

show abstract

“…Several temporal features have been used for script recognition in general and for on-line Devanagari script recognition in particular [22,14,33,11]. We discuss feature set based on directional properties of the curve connecting two…”

Section: Direction Based Feature Extractionmentioning

confidence: 99%

A Framework for On-Line Devanagari Handwritten Character Recognition

Kopparapu¹,

L²

2014

Preprint

View full text Add to dashboard Cite

The main challenge in on-line handwritten character recognition in Indian language is the large size of the character set, larger similarity between different characters in the script and the huge variation in writing style. In this paper we propose a framework for on-line handwitten script recognition taking cues from speech signal processing literature. The framework is based on identifying strokes, which in turn lead to recognition of handwritten on-line characters rather that the conventional character identification. Though the framework is described for Devanagari script, the framework is general and can be applied to any language.The proposed platform consists of pre-processing, feature extraction, recognition and post processing like the conventional character recognition but applied to strokes. The on-line Devanagari character recognition reduces to one of recognizing one of 69 primitives and recognition of a character is performed by recognizing a sequence of such primitives. We further show the impact of noise removal on on-line raw data which is usually noisy. The use of Fuzzy Directional Features to enhance the accuracy of stroke recognition is also described. The recognition results are compared with commonly used directional features in literature using several classifiers.

show abstract

Maximum mutual information training for an online neural predictive handwritten word recognition system

Cited by 7 publications

References 20 publications

Recognition of Online Turkish Handwriting using Transfer Learning

Recognition of Online Turkish Handwriting using Transfer Learning

Large vocabulary recognition for online Turkish handwriting with sublexical units

A Framework for On-Line Devanagari Handwritten Character Recognition

Contact Info

Product

Resources

About