Handwritten Brazilian Month Recognition: An Analysis of Two NN Architectures and a Rejection Mechanism

Kapp, Marcelo N.; Freitas, Cinthia Obladen de Almendra; Sabourin, Robert

doi:10.1109/iwfhr.2004.53

Cited by 9 publications

(6 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…[12]. Many researchers have used the Hidden-Markov Model (HMM), Neural-Networks (NN) and Multi-Layer Perceptrons (MLP), which provides accuracy, but involve lot of complexity [13,14]. The method applied in our experiment does not involve the complex structures used in previous work.…”

Section: Previous Workmentioning

confidence: 99%

A Projection Based Statistical Approach for Handwritten Character Recognition

Raju¹,

Satyanarayana²

2007

International Conference on Computational Intelligence and Multimedia Applications (ICCIMA 2007)

View full text Add to dashboard Cite

The aim of this paper is to recognize upper-case English alphabets from handwritten documents. An OCR system based on four-sided projections of an alphabet is proposed. The projection points are approximated by polygons and feature points are extracted subsequently for notch elimination and segmentation of the polygon. The resulting segments are smoothed by Bézier approximation. Statistical matching technique is successfully applied for character recognition.

show abstract

Section: Previous Workmentioning

confidence: 99%

A Projection Based Statistical Approach for Handwritten Character Recognition

Raju¹,

Satyanarayana²

2007

International Conference on Computational Intelligence and Multimedia Applications (ICCIMA 2007)

View full text Add to dashboard Cite

show abstract

“…ANNs (artificial neural networks) and HMMs (hidden markov models) are the most used, amongst the techniques which have been investigated for handwriting recognition. It has been observed that ANNs in general obtained best results than HMMs [11]. The most widely studied and used artificial neural network is the MLP (multi-layer perceptron) [12].…”

Section: Introductionmentioning

confidence: 99%

Offline Handwritten Characters Recognition Using Moments Features and Neural Networks

Abaynarh¹,

Zenkouar²

2015

CTA

View full text Add to dashboard Cite

In this paper we revise the moment theory for pattern recognition designed, to extract patterns from the noisy character datas, and develop unconstrained handwritten. Amazigh character recognition method based upon orthogonal moments and neural networks classifier. We argue that, given the natural flexibility of neural network models and the extent of parallel processing that they allow, our algorithm is a step forward in character recognition. More importantly, following the approach proposed, we apply our system to two different databases, to examine the ability to recognize patterns under noise. We discover overwhelming support for different style of writing. Moreover, this basic conclusion appears to remain valid across different levels of smoothing and insensitive to the nuances of character patterns. Experiments tested the effect of set size on recognition accuracy which can reach 97.46%. The novelty of the proposed method is independence of size, slant, orientation, and translation. The performance of the proposed method is experimentally evaluated and the promising results and findings are presented. Our method is compared to K-NN (k-nearest neighbors) classifier algorithm; results show performances of our method.

show abstract

“…In the literature we can find a wide diversity of VSs for HWR. On the one hand are the VSs directly applying a rejection rule to the HWR hypotheses scores [8,7,10]. For HWR based on Hidden Markov Models (HMMs), which is by far the most successfully employed statistical approach according to the state-of-theart, VS rejection mechanisms rely usually on the same HMM decoding scores.…”

Section: Introductionmentioning

confidence: 99%

Handwritten Word Verification by SVM-Based Hypotheses Re-scoring and Multiple Thresholds Rejection

Guichard

Toselli

Coüasnon

2010

2010 12th International Conference on Frontiers in Handwriting Recognition

View full text Add to dashboard Cite

In the field of isolated handwritten word recognition, the development of verification systems that optimize the trade-off between performance and reliability is still an active research topic. To minimize the recognition errors, usually, a verification system is used to accept or reject the hypotheses output by an existing recognition system. In this paper, a novel verification architecture is presented. In essence, the recognition hypotheses, re-scored by a set of the support vector machines, are validated by a verification mechanism based on multiple rejection thresholds. In order to tune these (classdependent) rejection thresholds, an algorithm based on dynamic programming is proposed which focus on maximizing the recognition rate for a given prefixed error rate. Preliminary reported results of experiments carried out on RIMES database show that this approach performs equal or superior to other state-of-the-art rejection methods.

show abstract

Handwritten Brazilian Month Recognition: An Analysis of Two NN Architectures and a Rejection Mechanism

Cited by 9 publications

References 7 publications

A Projection Based Statistical Approach for Handwritten Character Recognition

A Projection Based Statistical Approach for Handwritten Character Recognition

Offline Handwritten Characters Recognition Using Moments Features and Neural Networks

Handwritten Word Verification by SVM-Based Hypotheses Re-scoring and Multiple Thresholds Rejection

Contact Info

Product

Resources

About