Handwritten Amharic Character Recognition System Using Convolutional Neural Networks

Abdurahman, Fetulhak

doi:10.12739/nwsa.2019.14.2.1a0433

Cited by 6 publications

(4 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recently there are some encouraging works emerging on Amharic character recognition applying deep learning techniques. The different authors emphasized on different types of documents including printed [2,9,13], ancient [20,11], and handwritten documents [14,13,1]. Accordingly the research efforts in this regard lack complementing each other and improving results based on a clear baseline.…”

Section: Related Workmentioning

confidence: 99%

Multi-script Handwritten Digit Recognition Using Multi-task Learning

Gondere,

Schmidt-Thieme,

Sharma

et al. 2021

Preprint

View full text Add to dashboard Cite

Handwritten digit recognition is one of the extensively studied area in machine learning. Apart from the wider research on handwritten digit recognition on MNIST dataset, there are many other research works on various script recognition. However, it is not very common for multi-script digit recognition which encourage the development of robust and multipurpose systems. Additionally working on multi-script digit recognition enables multi-task learning, considering the script classification as a related task for instance. It is evident that multi-task learning improves model performance through inductive transfer using the information contained in related tasks. Therefore, in this study multi-script handwritten digit recognition using multi-task learning will be investigated. As a specific case of demonstrating the solution to the problem, Amharic handwritten character recognition will also be experimented. The handwritten digits of three scripts including Latin, Arabic and Kannada are studied to show that multi-task models with reformulation of the individual tasks have shown promising results. In this study a novel way of using the individual tasks predictions was proposed to help classification performance and regularize the different loss for the purpose of the main task. This finding has outperformed the baseline and the conventional multi-task learning models. More importantly, it avoided the need for weighting the different losses of the tasks, which is one of the challenges in multi-task learning.

show abstract

Section: Related Workmentioning

confidence: 99%

Multi-script Handwritten Digit Recognition Using Multi-task Learning

Gondere,

Schmidt-Thieme,

Sharma

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…In particular, the original ResNet-18 yielded 94.52% classification accuracy, while the modified ResNet-18 achieved 95.10%. In [14], a new CNN based system was proposed for Amharic character recognition. The dataset used by the authors includes 132,500 Amharic characters.…”

Section: Slant Correctionmentioning

confidence: 99%

Deep ensemble neural networks for recognizing isolated Arabic handwritten characters

Alyahya¹,

Ismail²,

Al-Salman³

2020

TIPCV

View full text Add to dashboard Cite

In recent years, handwritten character recognition has become an active research field. In particular, digitalization has triggered the interest of researchers from various computing disciplines to address several handwriting related challenges. Despite these efforts, there are still opportunities for the development and improvement of the recognition of the handwritten Arabic letters. In this paper, we designed and developed a deep ensemble architecture in which ResNet-18 architecture is exploited to model and classify character images. Specifically, we adapted ResNet-18 by adding a dropout layer after all convolutional layer and integrated it in multiple ensemble models to automatically recognize isolated handwritten Arabic characters. A standard Arabic Handwritten Character Dataset (AHCD) was used in the experiments to train and assess all the proposed models. Satisfactory results were obtained using all models. The best-attained accuracy was 98.30% using a typical ResNet-18 model. Similarly, 98.00% and 98.03% accuracies were obtained using an ensemble model with one fully connected layer (1 FC) and an ensemble with two fully connected layers (2 FC) coupled with a dropout layer, respectively.

show abstract

“…Assabie and Bigün [1] used a standard hand-engineered feature to represent character symbols and used HMM model to recognize handwritten words. The other related works considered isolated Amharic handwritten character recognition [23][24][25][26]. Recent studies in [27,28] used LSTM-RNNs in combination with CTC algorithms for an end-to-end Amharic Optical Character Recognition(OCR).…”

Section: Introductionmentioning

confidence: 99%

AHWR-Net: offline handwritten amharic word recognition using convolutional recurrent neural network

2021

Self Cite

View full text Add to dashboard Cite

Amharic ("Image missing") is the official language of the Federal Government of Ethiopia, with more than 27 million speakers. It uses an Ethiopic script, which has 238 core and 27 labialized characters. It is a low-resourced language, and a few attempts have been made so far for its handwritten text recognition. However, Amharic handwritten text recognition is challenging due to the very high similarity between characters. This paper presents a convolutional recurrent neural networks based offline handwritten Amharic word recognition system. The proposed framework comprises convolutional neural networks (CNNs) for feature extraction from input word images, recurrent neural network (RNNs) for sequence encoding, and connectionist temporal classification as a loss function. We designed a custom CNN model and compared its performance with three different state-of-the-art CNN models, including DenseNet-121, ResNet-50 and VGG-19 after modifying their architectures to fit our problem domain, for robust feature extraction from handwritten Amharic word images. We have conducted detailed experiments with different CNN and RNN architectures, input word image sizes, and applied data augmentation techniques to enhance performance of the proposed models. We have prepared a handwritten Amharic word dataset, HARD-I, which is available publicly for researchers. From the experiments on various recognition models using our dataset, a WER of 5.24 % and CER of 1.15 % were achieved using our best-performing recognition model. The proposed models achieve a competitive performance compared to existing models for offline handwritten Amharic word recognition.

show abstract

Handwritten Amharic Character Recognition System Using Convolutional Neural Networks

Cited by 6 publications

References 16 publications

Multi-script Handwritten Digit Recognition Using Multi-task Learning

Multi-script Handwritten Digit Recognition Using Multi-task Learning

Deep ensemble neural networks for recognizing isolated Arabic handwritten characters

AHWR-Net: offline handwritten amharic word recognition using convolutional recurrent neural network

Contact Info

Product

Resources

About