Handwritten Bangla word recognition using negative refraction based shape transformation

Sahoo, Samanway; Nandi, Subham Kumar; Barua, Sourav; Pallavi, .; Bhowmik, Showmik; Malakar, Samir; Sarkar, Ram

doi:10.3233/jifs-169712

Cited by 15 publications

(5 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The model requires a reliable ground truth or labeled data, which is difficult to handle. Sahoo et al [19] followed a holistic word recognition approach to recognize handwritten Bangla words. their dataset consists of 12000 handwritten Bangla words.…”

Section: Literature Reviewmentioning

confidence: 99%

Handwritten Word Recognition using Deep Learning Approach: A Novel Way of Generating Handwritten Words

Akter

Shahriar

Cuzzocrea

et al. 2022

2022 IEEE International Conference on Big Data (Big Data)

View full text Add to dashboard Cite

A handwritten word recognition system comes with issues such as-lack of large and diverse datasets. It is necessary to resolve such issues since millions of official documents can be digitized by training deep learning models using a large and diverse dataset. Due to the lack of data availability, the trained model does not give the expected result. Thus, it has a high chance of showing poor results. This paper proposes a novel way of generating diverse handwritten word images using handwritten characters. The idea of our project is to train the BiLSTM-CTC architecture with generated synthetic handwritten words. The whole approach shows the process of generating two types of large and diverse handwritten word datasets: overlapped and non-overlapped. Since handwritten words also have issues like overlapping between two characters, we have tried to put it into our experimental part. We have also demonstrated the process of recognizing handwritten documents using the deep learning model. For the experiments, we have targeted the Bangla language, which lacks the handwritten word dataset, and can be followed for any language. Our approach is less complex and less costly than traditional GAN models. Finally, we have evaluated our model using Word Error Rate (WER), accuracy, f1-score, precision, and recall metrics. The model gives 39% WER score, 92% percent accuracy, and 92% percent f1 scores using nonoverlapped data and 63% percent WER score, 83% percent accuracy, and 85% percent f1 scores using overlapped data.

show abstract

Section: Literature Reviewmentioning

confidence: 99%

Handwritten Word Recognition using Deep Learning Approach: A Novel Way of Generating Handwritten Words

Akter

Shahriar

Cuzzocrea

et al. 2022

2022 IEEE International Conference on Big Data (Big Data)

View full text Add to dashboard Cite

show abstract

“…For evaluation purposes, the author considers 18,000 samples of bangle handwritten word images that include roughly 120 city names where the prediction accuracy is 84% approximately. Similarly, Sahoo et al [4] discuss negative refraction properties with shape-feature descriptors to recognize the accuracy holistically. It includes 80 handwritten city names over bangle scripts.…”

Section: Literature Reviewmentioning

confidence: 99%

“…feature extraction approaches rely on naturally inspired heuristic models to produce potential features for classification purposes [4]. Thus, the optimal selections of features are complex to predict due to their varying nature.…”

mentioning

confidence: 99%

An efficient CNN model with squirrel optimizer for handwritten digit recognition

Senthil¹,

Rajan²,

Deepika³

2021

IJATEE

View full text Add to dashboard Cite

The modern way of handwritten digit recognition has various potential real-time applications like management of students' score sheets, reading the tax form, postal code reading, and cheque processing [1]. Generally, diverse numeral strings are made of overlapping, segregated, connected, or wrecked digits. Segmentation, feature selection, and feature extraction of digit handwritten images with diverse numeral string lengths are analyzed by various researchers [2]. *Author for correspondence Those existing approaches are designed to predict the unconstrained handwritten digit which includes all the four pre-processing stages before classification.The most challenging tasks with this prediction model are segmentation and feature extraction as these steps pretend to give better classification accuracy of isolated digits [3].Generally, the numbers of digits over, the strings are unknown and therefore the optimal boundary between them is extremely complex to identify. Some existing Re search Article

show abstract

“…To evaluate the scheme, they have prepared a database comprising 18,000 handwritten Bangla word images of 120 different city names and have obtained a maximum accuracy of 83.64%. While Sahoo et al [14] have lately proposed a shape‐based feature descriptor based on the negative refraction property of the light to holistically recognise 80 handwritten city names written in Bangla script. They have yielded an accuracy of 87.50% over 12,000 handwritten word images by combining both SMO and simple logistics classifiers.…”

Section: Related Workmentioning

confidence: 99%

“…Over the years, plenty of works have been reported for handwritten word recognition in Chinese/Japanese [5–7], Arabic [3, 8–10], and Roman scripts [11, 12]. However, in contrast, limited works [4, 13, 14] have been published for handwritten word recognition in Bangla script. Bangla is one of the most popular official languages of India and also the national language of Bangladesh.…”

Section: Introductionmentioning

confidence: 99%

H‐WordNet: a holistic convolutional neural network approach for handwritten word recognition

Das

Nayak

Dash

et al. 2020

IET Image Processing

View full text Add to dashboard Cite

Handwritten Bangla word recognition using negative refraction based shape transformation

Cited by 15 publications

References 29 publications

Handwritten Word Recognition using Deep Learning Approach: A Novel Way of Generating Handwritten Words

Handwritten Word Recognition using Deep Learning Approach: A Novel Way of Generating Handwritten Words

An efficient CNN model with squirrel optimizer for handwritten digit recognition

H‐WordNet: a holistic convolutional neural network approach for handwritten word recognition

Contact Info

Product

Resources

About