Debajit Datta | Vellore Institute of Technology (original) (raw)

Papers by Debajit Datta

Research paper thumbnail of Eye Gaze Detection Based on Computational Visual Perception and Facial Landmarks

Computers, Materials & Continua, 2021

Research paper thumbnail of Optimization of an Automated Examination Generation System Using Hybrid Recurrent Neural Network

Computing Methodology eJournal, 2020

Globally, there has been a drastic increase in student-teacher ratio, over the years, which has l... more Globally, there has been a drastic increase in student-teacher ratio, over the years, which has led to the popularity of the digital examination system over classrooms, which is also eco-friendly. As a result, various new challenges have occurred due to the digitization of online education systems – for those who are not well versed with the technology, and for tracking malpractices. This issue can be acknowledged by automated examination question generation that will generate a unique exam. Using Recurrent Neural Network (RNN) models, the questions can be generated, however, the commonly used models – like, the simple RNN model, the bidirectional RNN model, the RNN model with embeddings and the encoder-decoder RNN model – do not provide high accuracies. This work proposes a hybridized RNN model that incorporates multiple layers of these commonly used RNN models, in a specific order, creating an optimized model that will be able to generate questions with higher accuracy. Additional...

Research paper thumbnail of Image Classification Using CNN With Multi-Core and Many-Core Architecture

Applications of Artificial Intelligence for Smart Technology, 2021

Image classification is a widely discussed topic in this era. It covers a vivid range of applicat... more Image classification is a widely discussed topic in this era. It covers a vivid range of application domains like from garbage classification applications to advanced fields of medical sciences. There have been several research works that have been done in the past and are also currently under research for coming up with better-optimized image classification techniques. However, the process of image classification turns out to be time-consuming. This work deals with the widely accepted FashionMNIST (modified national institute of standards and technology database) dataset, having a set of sixty thousand images for training a model and another popular dataset of MNIST for handwritten numbers. The work compares several convolutional neural network (CNN) models and aims in parallelizing them using a distributed framework that is provided by the python library, RAY. The parallelization has been achieved over the multiple cores of CPU and many cores of GPU. The work also shows that the o...

Research paper thumbnail of Comparison of Performance of Parallel Computation of CPU Cores on CNN model

2020 International Conference on Emerging Trends in Information Technology and Engineering (ic-ETITE), 2020

Algorithms such as Convolutional Neural Network used for image classification. When huge convolut... more Algorithms such as Convolutional Neural Network used for image classification. When huge convolutional neural networks are used to classify image datasets it takes a lot of time to perform convolution operations, which results in increasing the computational demand of image processing. The better way to accelerate image processing is to parallelizing multiple CPU cores. CPU consists of a few numbers of cores that are optimized for sequential processing. 1GHz. This project aims to compare the performance of parallelized CPUs. To parallelize multicore CPUs Python’s Ray library is being used. Convolutional Neural Network is used as the benchmark image classification algorithm in this project. We have tried to show the speedup in training a model with parallelization of the multicores of a CPU. The comparative study in our work has been carried out amongst three different Convolutional Neural Network models namely, the AlexNet model, the VGG16 model and the model proposed by ‘emmarex’ which we found on the online deep-learning platform, called Kaggle. The dataset used in this project is Plant Disease Image Dataset.

Research paper thumbnail of An Efficient Sound and Data Steganography Based Secure Authentication System

Computers, Materials & Continua, 2021

Research paper thumbnail of Exploration of Various Attacks and Security Measures Related to the Internet of Things

Regular, 2020

In this era of technological advances, it will be impractical to think of a day without the usage... more In this era of technological advances, it will be impractical to think of a day without the usage of gadgets. Development and popularity of the Internet of Things have helped mankind a lot in several ways, but at the same time, there has also been an increase in attacks invading the underlying security. Advances in studies have resulted in the development of evolved algorithms that can be used in order to reduce the attacks and threats to the Internet of Things. With several advancements in studies and research works, the security measures on various Internet of Things based components and protocols are developing with time, but concurrently more advanced threats and attacks on these components are also evolving. These attacks are not only harmful to the components, but rather they also affect the users and applications that are associated with it, by breaching data, increase in inconsistency and inaccuracy, and many more. This work deals with the study of several attacks that are a...

Research paper thumbnail of Neural Machine Translation using Recurrent Neural Network

International Journal of Engineering and Advanced Technology, 2020

In this era of globalization, it is quite likely to come across people or community who do not sh... more In this era of globalization, it is quite likely to come across people or community who do not share the same language for communication as us. To acknowledge the problems caused by this, we have machine translation systems being developed. Developers of several reputed organizations like Google LLC, have been working to bring algorithms to support machine translations using machine learning algorithms like Artificial Neural Network (ANN) in order to facilitate machine translation. Several Neural Machine Translations have been developed in this regard, but Recurrent Neural Network (RNN), on the other hand, has not grown much in this field. In our work, we have tried to bring RNN in the field of machine translations, in order to acknowledge the benefits of RNN over ANN. The results show how RNN is able to perform machine translations with proper accuracy.

Research paper thumbnail of Performance Comparison of Deep CNN Models for Detecting Driver’s Distraction

Computers, Materials & Continua, 2021

Research paper thumbnail of Eye Gaze Detection Based on Computational Visual Perception and Facial Landmarks

Computers, Materials & Continua, 2021

Research paper thumbnail of Optimization of an Automated Examination Generation System Using Hybrid Recurrent Neural Network

Computing Methodology eJournal, 2020

Globally, there has been a drastic increase in student-teacher ratio, over the years, which has l... more Globally, there has been a drastic increase in student-teacher ratio, over the years, which has led to the popularity of the digital examination system over classrooms, which is also eco-friendly. As a result, various new challenges have occurred due to the digitization of online education systems – for those who are not well versed with the technology, and for tracking malpractices. This issue can be acknowledged by automated examination question generation that will generate a unique exam. Using Recurrent Neural Network (RNN) models, the questions can be generated, however, the commonly used models – like, the simple RNN model, the bidirectional RNN model, the RNN model with embeddings and the encoder-decoder RNN model – do not provide high accuracies. This work proposes a hybridized RNN model that incorporates multiple layers of these commonly used RNN models, in a specific order, creating an optimized model that will be able to generate questions with higher accuracy. Additional...

Research paper thumbnail of Image Classification Using CNN With Multi-Core and Many-Core Architecture

Applications of Artificial Intelligence for Smart Technology, 2021

Image classification is a widely discussed topic in this era. It covers a vivid range of applicat... more Image classification is a widely discussed topic in this era. It covers a vivid range of application domains like from garbage classification applications to advanced fields of medical sciences. There have been several research works that have been done in the past and are also currently under research for coming up with better-optimized image classification techniques. However, the process of image classification turns out to be time-consuming. This work deals with the widely accepted FashionMNIST (modified national institute of standards and technology database) dataset, having a set of sixty thousand images for training a model and another popular dataset of MNIST for handwritten numbers. The work compares several convolutional neural network (CNN) models and aims in parallelizing them using a distributed framework that is provided by the python library, RAY. The parallelization has been achieved over the multiple cores of CPU and many cores of GPU. The work also shows that the o...

Research paper thumbnail of Comparison of Performance of Parallel Computation of CPU Cores on CNN model

2020 International Conference on Emerging Trends in Information Technology and Engineering (ic-ETITE), 2020

Algorithms such as Convolutional Neural Network used for image classification. When huge convolut... more Algorithms such as Convolutional Neural Network used for image classification. When huge convolutional neural networks are used to classify image datasets it takes a lot of time to perform convolution operations, which results in increasing the computational demand of image processing. The better way to accelerate image processing is to parallelizing multiple CPU cores. CPU consists of a few numbers of cores that are optimized for sequential processing. 1GHz. This project aims to compare the performance of parallelized CPUs. To parallelize multicore CPUs Python’s Ray library is being used. Convolutional Neural Network is used as the benchmark image classification algorithm in this project. We have tried to show the speedup in training a model with parallelization of the multicores of a CPU. The comparative study in our work has been carried out amongst three different Convolutional Neural Network models namely, the AlexNet model, the VGG16 model and the model proposed by ‘emmarex’ which we found on the online deep-learning platform, called Kaggle. The dataset used in this project is Plant Disease Image Dataset.

Research paper thumbnail of An Efficient Sound and Data Steganography Based Secure Authentication System

Computers, Materials & Continua, 2021

Research paper thumbnail of Exploration of Various Attacks and Security Measures Related to the Internet of Things

Regular, 2020

In this era of technological advances, it will be impractical to think of a day without the usage... more In this era of technological advances, it will be impractical to think of a day without the usage of gadgets. Development and popularity of the Internet of Things have helped mankind a lot in several ways, but at the same time, there has also been an increase in attacks invading the underlying security. Advances in studies have resulted in the development of evolved algorithms that can be used in order to reduce the attacks and threats to the Internet of Things. With several advancements in studies and research works, the security measures on various Internet of Things based components and protocols are developing with time, but concurrently more advanced threats and attacks on these components are also evolving. These attacks are not only harmful to the components, but rather they also affect the users and applications that are associated with it, by breaching data, increase in inconsistency and inaccuracy, and many more. This work deals with the study of several attacks that are a...

Research paper thumbnail of Neural Machine Translation using Recurrent Neural Network

International Journal of Engineering and Advanced Technology, 2020

In this era of globalization, it is quite likely to come across people or community who do not sh... more In this era of globalization, it is quite likely to come across people or community who do not share the same language for communication as us. To acknowledge the problems caused by this, we have machine translation systems being developed. Developers of several reputed organizations like Google LLC, have been working to bring algorithms to support machine translations using machine learning algorithms like Artificial Neural Network (ANN) in order to facilitate machine translation. Several Neural Machine Translations have been developed in this regard, but Recurrent Neural Network (RNN), on the other hand, has not grown much in this field. In our work, we have tried to bring RNN in the field of machine translations, in order to acknowledge the benefits of RNN over ANN. The results show how RNN is able to perform machine translations with proper accuracy.

Research paper thumbnail of Performance Comparison of Deep CNN Models for Detecting Driver’s Distraction

Computers, Materials & Continua, 2021