Multilingual handwritten numeral recognition using a robust deep network joint with transfer learning

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 state-of-the-art datasets and approaches towards offline BHDR have been analyzed. Furthermore, several real-life application-specific 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.

https://ieeexplore.ieee.org/ielx7/6287639/6514899/09869842.pdf

Two Decades of Bengali Handwritten Digit Recognition: A Survey

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 application-specific 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. 

/pdf/two-decades-of-bengali-handwritten-digit-recognition-a-7jcj0924.pdf

Recognition of handwritten digits is a stimulating task in recent years. Even though many deep learning-oriented classification algorithms are deliberated for handwritten digit recognition, the accuracy and time still want to be more improved. It is very normal to deal with a issue known as premature conjunction. This problematic property is widely considered in swarm intelligence algorithms such as Particle Swarm Optimisation (PSO). In order to overwhelm the deficiencies of traditional PSO, a novel model is intended with Convolutional Neural Networks (CNN). CNN is built by modifying the neural network assembly with Serendipity and Orthogonal Learning Particle Swarm Optimization (CNN-SOLPSO). This alteration is received for developmentally advancing the quantity of hyper-boundaries. This projected enhancer expects the ideal qualities from that wellness calculation and illustrates enhanced productivity while contrasted with different extra customary methodologies. The MNIST dataset of transcribed digits is utilized for preparing and testing the calculation embraced in the proposed model. The MNIST dataset comprises of manually written digits pictures which are assorted and profoundly contorted. A definitive objective of this work is to attempt an appropriate way approach digitalization by contribution predominant exactness and improved calculation. MATLAB 2018b is used when the replication environment to determines parameter like Validation exactness and loss measurements, Training exactness and loss capacity and Recognition velocity with fault rate and completing moment.

Handwritten Digits Recognition from Images using Serendipity and Orthogonal Schemes

Identification of numbers has gained excitement recently. Despite the fact that several learning focused categorization approaches are proposed for mnist dataset validation, the precision and processing time may still be improved. Dealing with a disease as an early union is rather common. Swarm approaches like Swarm Optimization seriously evaluate this unfavorable element (PSO). A novel approach using neural network models with convolutions is intended to address the limitations of traditional Soc (CNN). Clo is created by modifying the artificial neural network with the use of luck and analogous learnt optimized particle swarms (CNN-SOLPSO). This adaption is provided for the steadily growing population of the over. This projected enhancer shows increased efficacy when compared to other unconventional methods and expects the best characteristics from that wellbeing assessment. The Holdout library of transcribed digits is used to construct and evaluate the computation contained in the proposed model. the severely deformed, unpredictable, and manually produced pictures of digits that help compensate its Imagenet dataset database. The major objective of this effort is to contribute to an appropriate approach to digital by focusing on greater precision and better computations. Using Bas 2018b, it is possible to choose parameters for Training unshakable quality and drop capacity, Validate refinement and loss measurements, and Identify velocities with defect rate and completion moment. 

Orthogonal Schemes for Handwritten Digits Recognizing from Image Data

Indian subcontinent is a birthplace of multilingual people, where documents such as job application form, passport, number plate identification, and so forth are composed of text contents written in different languages or scripts. These scripts consist of different Indic numerals in a single document page. Recently, deep convolutional neural networks (CNN) have achieved favorable result in computer vision problems, especially in recognizing handwritten digits but most of the works focuses on only one language, i.e., English or Hindi or Bangla, etc. However, developing a language-invariant method is very important as we live in a global village now. In this work, we have examined the performance of the ten state-of-the-art deep CNN methods for the recognition of handwritten digits using four most common languages in the Indian sub-continent that creates the foundation of a script invariant handwritten digit recognition system. Among the deep CNNs, Inception-v4 performs the best based on accuracy and computation time. Besides, it discusses the limitations of existing techniques and shows future research directions.

Sarwar Mahmud Milon

Papers

Quantitative Analysis of Deep CNNs for Multilingual Handwritten Digit Recognition