Recognition of Pashto Handwritten Characters Based on Deep Learning

Abstract

Handwritten character recognition is increasingly important in a variety of automation fields, for example, authentication of bank signatures, identification of ZIP codes on letter addresses, and forensic evidence. Despite improved object recognition technologies, Pashto’s hand-written character recognition (PHCR) remains largely unsolved due to the presence of many enigmatic hand-written characters, enormously cursive Pashto characters, and lack of research attention. We propose a convolutional neural network (CNN) model for recognition of Pashto hand-written characters for the first time in an unrestricted environment. Firstly, a novel Pashto handwritten character data set, “Poha”, for 44 characters is constructed. For preprocessing, deep fusion image processing techniques and noise reduction for text optimization are applied. A CNN model optimized in the number of convolutional layers and their parameters outperformed common deep models in terms of accuracy. Moreover, a set of benchmark popular CNN models applied to Poha is evaluated and compared with the proposed model. The obtained experimental results show that the proposed model is superior to other models with test accuracy of 99.64 percent for PHCR. The results indicate that our model may be a strong candidate for handwritten character recognition and automated PHCR applications.

Sports Deep Learning Method Based on Cognitive Human Behavior Recognition

“…is method can effectively reduce the dependence of data on shooting angles. Amin et al first selected four bone key points, and then calculated the geometric vector between all other bone joint points and the four key joint points in the cylindrical coordinate system as the frame-level feature vector [17].…”

“…Receive the 1D feature vector output by full connection, and then use Softmax classifier to complete the mapping from eigenvalue to probability. If the category of the classification task is K, Softmax maps the input Y ∈ R n to K (0, 1) real numbers and ensures that their sum is 1, as shown in equation (17). Assuming that the output is P ∈ R k , the mapping expression is shown in the following equation.…”

Human Sentiment and Activity Recognition in Disaster Situations Using Social Media Images Based on Deep Learning

“…In the case of human activity recognition, the deep models require large training data; to tackle this problem, the transfer learning approach has been thoroughly studied [29]. Since many promising outcomes have been obtained, a widely agreed issue is that it is very costly to annotate all the human activities, as human annotators and deep learning experts offer numerous efforts, and the probability of error still remains high [30,31]. Nevertheless, providing reliable and Sensors 2020, 20, 7115 5 of 25 relevant details is a salient task in human physical activity detection.…”

PHTI: Pashto Handwritten Text Imagebase for Deep Learning Applications

“…In addition to that, the development of datasets becomes more challenging while it deals with a language whose research is just in its introductory phase [7]. Further, the researchers always try to find out an appropriate dataset that covers all the possible levels of diversity in the target language [12], [13]. Handwritten text recognition has become an important area in the DIA and finding a good dataset for recognition is the main issue.…”