Article
Chemistry, Multidisciplinary
Amin Alqudah, Ali Mohammad Alqudah, Hiam Alquran, Hussein R. Al-Zoubi, Mohammed Al-Qodah, Mahmood A. Al-Khassaweneh
Summary: This paper proposes a two-stage methodology for the detection and classification of Arabic and Hindi handwritten numerals using convolutional neural networks (CNNs). The simulation results demonstrate high accuracy and performance, with a recognition rate close to 100%.
APPLIED SCIENCES-BASEL
(2021)
Article
Computer Science, Information Systems
Abu Sufian, Anirudha Ghosh, Avijit Naskar, Farhana Sultana, Jaya Sil, M. M. Hafizur Rahman
Summary: This paper presents a Bengali handwritten numeral digit recognition model based on densely connected convolutional neural networks (BDNet), which achieves high accuracy on the test dataset and significantly reduces errors compared to state-of-the-art models. The authors also create a dataset of Bengali handwritten numeral images for testing the trained model.
JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES
(2022)
Article
Computer Science, Artificial Intelligence
Rami S. Alkhawaldeh
Summary: This research introduces a deep hybrid transfer model for Arabic (Indian) handwritten digit recognition, utilizing CNN and LSTM for feature learning, achieving significant performance improvement.
Article
Computer Science, Artificial Intelligence
Mir Moynuddin Ahmed Shibly, Tahmina Akter Tisha, Tanzina Akter Tani, Shamim Ripon
Summary: This study aims to classify handwritten Bangla characters through three phases, utilizing convolutional neural networks and ensemble methods to achieve higher performance, ultimately achieving accuracy, precision, and recall of 98.68%, 98.69%, and 98.68% respectively.
PEERJ COMPUTER SCIENCE
(2021)
Article
Computer Science, Artificial Intelligence
Huseyin Kusetogullari, Amir Yavariabdi, Johan Hall, Niklas Lavesson
Summary: This paper presents the DIGITNET deep learning architecture and DIDA digit dataset for detecting and recognizing digits in historical handwritten documents from the nineteenth century. The dataset is generated from 100,000 Swedish historical document images and contains three sub-datasets for training the DIGITNET network, which outperforms existing methods according to experimental results.
Article
Engineering, Electrical & Electronic
Abdulrahman Saad Alqahtani, A. Neela Madheswari, Azath Mubarakali, P. Parthasarathy
Summary: This paper discusses the applications of machine learning and its extended field, deep learning, in handwritten digit recognition. It proposes various deep learning algorithms and conducts design and analysis using the MNIST dataset.
OPTICAL AND QUANTUM ELECTRONICS
(2023)
Article
Computer Science, Information Systems
Mohammad Meraj Khan, Mohammad Shorif Uddin, Mohammad Zavid Parvez, Lutfur Nahar
Summary: This paper presents a deep CNN model using SE-ResNeXt to recognize Bangla handwritten compound characters, achieving an average accuracy of 99.82% and outperforming state-of-the-art models in the experiment.
JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES
(2022)
Article
Computer Science, Artificial Intelligence
Md. Nahidul Islam Opu, Md. Ekramul Hossain, Muhammad Ashad Kabir
Summary: This study evaluates the state-of-the-art deep learning models for handwritten Bangla character recognition (HBCR) and proposes a new lightweight model. The experimental results show that the proposed model outperforms other models in terms of efficiency and model size, while maintaining competitive accuracy.
NEURAL COMPUTING & APPLICATIONS
(2023)
Article
Computer Science, Information Systems
Parth Goel, Amit Ganatra
Summary: This study addresses the problem of handwritten Gujarati numeral recognition in India. It uses deep transfer learning to find the best performing model by freezing and fine-tuning the weight parameters of ten pre-trained CNN architectures.
Article
Computer Science, Artificial Intelligence
Rami S. Alkhawaldeh, Moatsum Alawida, Nawaf Farhan Funkur Alshdaifat, Wafa' Za'al Alma'aitah, Ammar Almasri
Summary: This study focuses on Arabic (Indian) digits and proposes an ensemble deep transfer learning (EDTL) model that effectively detects and recognizes these digits. The EDTL model is a combination of two effective pre-trained transfer learning models, with time and cost complexity in the training phase.
NEURAL COMPUTING & APPLICATIONS
(2022)
Article
Chemistry, Multidisciplinary
Md. Humaun Kabir, Faruk Ahmad, Md. Al Mehedi Hasan, Jungpil Shin
Summary: Individuals' names have significant meaning and gender differences. This research focuses on gender prediction based on Bengali names using DL-based methods. Experimental results show that Conv1D achieves the highest accuracy of 91.18%.
APPLIED SCIENCES-BASEL
(2023)
Article
Computer Science, Artificial Intelligence
Rahul Pramanik, Soumen Bag
Summary: This research focuses on the decomposition problem in recognizing Bangla cursive words, achieving 98.86% accuracy by utilizing different feature sets and classifiers. The study suggests that the segmentation-free approach performs well in handling holistic vocabulary and merits further investigation.
NEURAL COMPUTING & APPLICATIONS
(2021)
Article
Computer Science, Artificial Intelligence
Abhishek Hazra, Prakash Choudhary, Sanasam Inunganbi, Mainak Adhikari
Summary: The paper introduces a unique CNN architecture for recognizing handwritten characters in Bangla and Meitei Mayek scripts. The model is validated on different datasets and shows excellent performance.
APPLIED INTELLIGENCE
(2021)
Article
Computer Science, Information Systems
Ruhul Amin, Md. Shamim Reza, Yuichi Okuyama, Yoichi Tomioka, Jungpil Shin
Summary: Recognition of Bengali handwritten digits is challenging due to variations in writing styles, shapes and sizes, noise levels, and image distortion. This paper introduces a new dataset and custom CNN models that outperform other models in terms of accuracy and resilience. The research provides important implications for various domains.
Article
Computer Science, Artificial Intelligence
Andre G. Hochuli, Alceu S. Britto Jr, David A. Saji, Jose M. Saavedra, Robert Sabourin, Luiz S. Oliveira
Summary: Traditional approaches for handwritten digit string recognition have relied on digit segmentation, while recent segmentation-free strategies offer a new perspective but still show limitations when dealing with a large number of touching digits. This study introduces an approach that treats a string of digits as a sequence of objects, evaluating various end-to-end methods to solve the HDSR problem.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)
Article
Computer Science, Artificial Intelligence
Bitanu Chatterjee, Trinav Bhattacharyya, Kushal Kanti Ghosh, Agneet Chatterjee, Ram Sarkar
Summary: This article presents a framework for maximizing influence propagation in a social network, which includes community detection and the utilization of the Shuffled Frog Leaping algorithm. Experimental results show that our method performs well compared to other algorithms.
Article
Computer Science, Information Systems
Soham Chattopadhyay, Arijit Dey, Pawan Kumar Singh, Ali Ahmadian, Ram Sarkar
Summary: Speech is crucial in human communication and human-computer interaction. In the field of AI and ML, it has been extensively studied to recognize human emotions from speech signals. To address the challenge of large feature dimension, a hybrid feature selection algorithm called CEOAS is proposed. By extracting LPC and LPCC features, the proposed model reduces feature dimension and improves classification accuracy. Impressive recognition accuracies have been achieved on four benchmark datasets, surpassing state-of-the-art algorithms.
MULTIMEDIA TOOLS AND APPLICATIONS
(2023)
Article
Computer Science, Artificial Intelligence
Ashis Paul, Arpan Basu, Mufti Mahmud, M. Shamim Kaiser, Ram Sarkar
Summary: This article discusses the use of deep learning models and an inverted bell-curve weighted ensemble method to assist in the detection of COVID-19 in CXR images. By using transfer learning and retraining models pretrained on the ImageNet dataset, as well as performing weighted average predictions, the accuracy of COVID-19 identification in CXR images can be improved.
NEURAL COMPUTING & APPLICATIONS
(2023)
Article
Computer Science, Software Engineering
Samir Malakar, Samanway Sahoo, Anuran Chakraborty, Ram Sarkar, Mita Nasipuri
Summary: Handwritten word recognition is an open research problem due to variations in writing style and degraded images. This paper proposes a holistic approach combined with distance calculation and feature descriptors to address the problem. The experimental results demonstrate the effectiveness of the proposed method on standard databases compared to deep learning models.
Article
Computer Science, Artificial Intelligence
Erik Cuevas, Hector Escobar, Ram Sarkar, Heba F. Eid
Summary: This paper proposes a new population initialization method for metaheuristic algorithms, where the initial set of candidate solutions is obtained through the sampling of the objective function. The method aims to find initial solutions that are close to the prominent values of the objective function, and these initial points represent promising regions of the search space. The proposed approach shows faster convergence and improved quality of solutions compared to other similar approaches.
APPLIED INTELLIGENCE
(2023)
Article
Computer Science, Artificial Intelligence
Rishav Pramanik, Payel Pramanik, Ram Sarkar
Summary: Breast cancer is a leading cause of premature death among women globally, but early detection and diagnosis can save lives. Hence, computer scientists are working to develop reliable models to tackle this disease. A proposed lightweight model combines transfer learning-based deep learning (DL) with feature selection to detect abnormalities in breast thermograms. This model performs well in detecting and differentiating malignant and healthy breasts.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Computer Science, Information Systems
Sagnik Ganguly, Sanmit Mandal, Samir Malakar, Ram Sarkar
Summary: This paper introduces a new copy-move image forgery detection technique which relies on a texture feature descriptor called Local Tetra Pattern (LTrP) for block level image comparison used to localize tampered region(s). Experimental results demonstrate that the proposed technique has been able to detect the forged regions with higher accuracy as compared to many state-of-the-art copy-move forgery detection methods.
MULTIMEDIA TOOLS AND APPLICATIONS
(2023)
Article
Computer Science, Artificial Intelligence
Payel Pramanik, Souradeep Mukhopadhyay, Seyedali Mirjalili, Ram Sarkar
Summary: Breast cancer is a common malignancy in women, and early detection is crucial. In this research, a method for classifying breast masses using mammograms is proposed. Deep features are extracted using the VGG16 model with an attention mechanism, and an optimal features subset is obtained using a meta-heuristic algorithm. The proposed model shows successful identification and differentiation of malignant and healthy breasts.
NEURAL COMPUTING & APPLICATIONS
(2023)
Correction
Computer Science, Artificial Intelligence
Apu Sarkar, S. K. Sabbir Hossain, Ram Sarkar
NEURAL COMPUTING & APPLICATIONS
(2023)
Article
Computer Science, Artificial Intelligence
Sk Mohiuddin, Khalid Hassan Sheikh, Samir Malakar, Juan D. Velasquez, Ram Sarkar
Summary: Digital face manipulation has become a significant concern recently due to its harmful effects on society, particularly for high-profile celebrities who can easily be targeted using apps like FaceSwap and FaceApp. Detecting deepfake images or videos is challenging, and existing models often fail to check for irrelevant or redundant features. In this study, a hierarchical feature selection (HFS) method using a hybrid population-based meta-heuristic model and a single solution-based meta-heuristic model was proposed. The model achieved high AUC scores on three publicly available datasets and outperformed most state-of-the-art methods.
NEURAL COMPUTING & APPLICATIONS
(2023)
Article
Computer Science, Artificial Intelligence
Samriddha Majumdar, Payel Pramanik, Ram Sarkar
Summary: Breast cancer is the second deadliest disease among women globally. Histopathology image analysis is an effective method for detecting tumor malignancies. Computer-aided diagnosis (CAD) using convolutional neural network (CNN) models has shown potential in breast histopathological image classification, but there is room for improvement. This paper proposes a novel rank-based ensemble method that combines multiple CNN models to enhance classification accuracy.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Computer Science, Information Systems
S. k Mohiuddin, Samir Malakar, Ram Sarkar
Summary: Video forgery has become more common due to the easy availability of tools. This study proposes an ensemble based method to detect duplicate frames in a video. By extracting different types of features and applying lexicographical sorting, the method achieves high detection accuracy and outperforms state-of-the-art methods.
MULTIMEDIA TOOLS AND APPLICATIONS
(2023)
Article
Computer Science, Information Systems
Sk Mohiuddin, Samir Malakar, Munish Kumar, Ram Sarkar
Summary: Video plays a critical role in conveying authenticity in various fields such as surveillance, medicine, journalism, and social media. However, the trust in videos is diminishing due to the ease of video forgery using accessible editing tools. This article comprehensively discusses the initiatives and recent trends in video forgery detection research worldwide.
MULTIMEDIA TOOLS AND APPLICATIONS
(2023)
Article
Computer Science, Information Systems
Avirup Bhattacharyya, Avigyan Bhattacharya, Sourajit Maity, Pawan Kumar Singh, Ram Sarkar
Summary: Designing an automatic vehicle detection system that caters to the requirements of the traffic management system is important. This research develops a still image database, JUVDsi v1, for designing an automated traffic management system in India. The database addresses the shortcomings of existing databases and is evaluated using state-of-the-art deep learning architectures.
MULTIMEDIA TOOLS AND APPLICATIONS
(2023)
Article
Computer Science, Interdisciplinary Applications
Ritam Guha, Kushal Kanti Ghosh, Suman Kumar Bera, Ram Sarkar, Seyedali Mirjalili
Summary: This paper proposes a binary adaptation of Equilibrium Optimizer (EO) called Discrete EO (DEO) for solving binary optimization problems. DEOSA algorithm, combining DEO with Simulated Annealing (SA) as a local search procedure, is applied to various datasets and outperforms other algorithms. The scalability and robustness of DEOSA are also tested on high-dimensional Microarray datasets and Knapsack problems, showing its superiority.
JOURNAL OF COMPUTATIONAL SCIENCE
(2023)