Handwritten Bangla Numeral Recognition Using Ensembling of Convolutional Neural Network

Noor, Rouhan; Islam, Kazi Mejbaul; Rahimi, Md. Jakaria

doi:10.1109/iccitechn.2018.8631944

Cited by 10 publications

(22 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Their ensemble model from their six best models, applied on NumtaDB dataset [25], achieved 99.3359% test accuracy. In [26] R. Noor et.al proposed an ensemble model based Convolutional Neural Network for recognizing Bengali handwritten numerals. They train their model in many noisy conditions using customized NumtaDB dataset [25].…”

Section: Existing Work On Bengali Handwritten Numeral Recognitionmentioning

confidence: 99%

BDNet: Bengali Handwritten Numeral Digit Recognition based on Densely connected Convolutional Neural Networks

Sufian

Ghosh

Naskar

et al. 2022

Journal of King Saud University - Computer and Information Scie

View full text Add to dashboard Cite

Bengali handwritten digit recognition can be done using different image classification techniques. But the images of handwritten digits are different from natural images as the orientation of a digit as well as similarity of features of different digits are important. On the other hand, deep convolutional neural networks are achieving huge success in computer vision problems, especially in image classification. This BDNet is a densely connected deep convolutional neural network model based on state-of-the-art algorithm DenseNet to classify Bengali handwritten numeral digits. The BDNet has end-to-end trained using ISI Bengali handwritten numeral dataset with 5-fold cross-validation. The BD-Net has achieved a test accuracy of 99.65%(baseline was 99.40%) on test data of ISI Bengali handwritten numerals. The trained model also gives 97.50% on own created dataset(which are not used during training). That is, this model gives a 41.66% error reduction compared to the previous state-of-the-art model. Codes, trained model and own dataset available at: https://github.com/Sufianlab/BDNet.

show abstract

Section: Existing Work On Bengali Handwritten Numeral Recognitionmentioning

confidence: 99%

BDNet: Bengali Handwritten Numeral Digit Recognition based on Densely connected Convolutional Neural Networks

Sufian

Ghosh

Naskar

et al. 2022

Journal of King Saud University - Computer and Information Scie

View full text Add to dashboard Cite

show abstract

“…Most of the existing literature on BHDR either use 28 × 28 [35]- [43] or 32 × 32 [44]- [59] input dimensions. Other dimensions include 16 × 16 [60], 48 × 48 [61], 64 × 64 [62], and 128 × 128 [28], [63]. The popularity of smaller input dimensions indicate that digit images can be reduced in size without affecting the performance of the learning model.…”

Section: ) Resizingmentioning

confidence: 99%

“…To avoid this, images are converted to binary form via thresholding techniques (such as Otsu's method [40], [54], [62], adaptive thresholding [47], [71]) and then inverted, so that the digits have an intensity value of 255 and the background has an intensity value of 0. This also helps get rid of random noises that might be present in the background of the digits.…”

Section: ) Color Inversionmentioning

confidence: 99%

“…9(d)). In BHDR, 0 to 30% zooming augmentation can be seen [56], [62], [63], [73], [77], [78]. Shifting requires translating each pixel of a sample by a constant factor chosen randomly within a certain range (Fig.…”

Section: A Geometric Transformation 1) Rotation and Flippingmentioning

confidence: 99%

“…The magnitude can also be controlled using a constant factor. Common values for constant factors are often in the range of 0 to 0.25 [61], [62], [77]. Shearing in this manner can help the model learn to recognize slanted texts.…”

Section: ) Distortionmentioning

confidence: 99%

See 2 more Smart Citations

Two Decades of Bengali Handwritten Digit Recognition: A Survey

et al. 2022

View full text Add to dashboard Cite

Handwritten Digit Recognition (HDR) is one of the most challenging tasks in the domain of Optical Character Recognition (OCR). Irrespective of language, there are some inherent challenges of HDR, which mostly arise due to the variations in writing styles across individuals, writing medium and environment, inability to maintain the same strokes while writing any digit repeatedly, etc. In addition to that, the structural complexities of the digits of a particular language may lead to ambiguous scenarios of HDR. Over the years, researchers have developed numerous offline and online HDR pipelines, where different image processing techniques are combined with traditional Machine Learning (ML)-based and/or Deep Learning (DL)-based architectures. Although evidence of extensive review studies on HDR exists in the literature for languages, such as English, Arabic, Indian, Farsi, Chinese, etc., few surveys on Bengali HDR (BHDR) can be found, which lack a comprehensive analysis of the challenges, the underlying recognition process, and possible future directions. In this paper, the characteristics and inherent ambiguities of Bengali handwritten digits along with a comprehensive insight of two decades of the state-of-the-art datasets and approaches towards offline BHDR have been analyzed. Furthermore, several real-life applicationspecific studies, which involve BHDR, have also been discussed in detail. This paper will also serve as a compendium for researchers interested in the science behind offline BHDR, instigating the exploration of newer avenues of relevant research that may further lead to better offline recognition of Bengali handwritten digits in different application areas.

show abstract