Saroj Kumar Chandra
@opju.ac.in
Assistant Professor Computer Science and Engineering
OP Jindal University Raigarh
I am Saroj Kumar Chandra, I have completed my Ph.D. (Computer Science and Engineering) from Indian Institute of Information Technology, Design and
Manufacturing, Jabalpur (M.P.), India in the year 2020. I have completed my M.Tech. (Computer Science and Engineering) degree from National Institute of Technology, Durgapur (W.B.), India in the year 2010 and My B.E. (Information Technology and Engineering) degree from Institute of Technology, Guru Ghasidas Central University, Bilaspur (C.G.), India in the year 2007. I have five- and half-year teaching experience as Assistant Professor. Also, I have four-year research experience as a Ph.D. scholar
EDUCATION
Computer Science and Engineering, Indian Institute of Information Technology, Design and Manufacturing, Jabalpur, Madhya-Pradesh, India, 2015- 2020
M.Tech.: Computer Science and Engineering, National Institute of Technology, Durgapur, West-Bengal, India, 2008-2010.
B.E.: Information Technology and Engineering, Institute of Technology, Guru Ghasidas Central University, Bilaspur, Chhattisgarh, India, 2002-2007.
RESEARCH INTERESTS
Image Processing
Machine Learning
Deep Learrning
Data Science
Block Chain
Scopus Publications
Scholar Citations
Scholar h-index
Scholar i10-index
Scopus Publications
Suman Patel, Saroj Kumar Chandra, and Amit Jain
IEEE
Deepfake videos, which are artificial intelligence-altered videos that have acquired extensive awareness recently. Deep learning algorithms are utilized to create these deepfake films, which are meant to disseminate false information about anyone, including politicians and celebrities. These movies have been purposefully made viral in order to propagate propaganda and false information, to frighten people, and to destabilize society. It is exceedingly challenging for a casual viewer to recognize a deep fake video with the naked eye. Finding these manipulated films has been extremely difficult and requires careful attention. In order to recognize a deepfake movie, this paper, a novel methodology that combines ResNext, a Convolutional Neural Network (CNN) algorithm, with Long Short-Term Memory (LSTM), a Recurrent Neural Network (RNN) has been developed. It is found that as the number of epochs increased, the model's accuracy increased and its training loss reduced.
Saroj Kumar Chandra
Chapman and Hall/CRC
Sourabh Kumar and Saroj Kumar Chandra
Springer Nature Singapore
Saroj Kumar Chandra, Ram Narayan Shukla, and Ashok Bhansali
Springer Nature Singapore
Sunny Singh and Saroj Kumar Chandra
Springer International Publishing
Uma Shankar Pandey, Gulshan Soni, and Saroj Kumar Chandra
IEEE
Sensor nodes work cooperatively to achieve a common objective in WSNs. These sensors are powered by battery cells. How to enhance the lifetime of the sensor nodes is still an area of research and several researchers attempted to enhance the lifetime of sensor nodes. The mechanism of the duty cycle provides a way to mitigate unnecessary power consumption in sensor networks. This paper presents a comprehensive survey of dynamic duty cycle schemes and a framework that provides a smart dynamic duty cycle mechanism for WSNs. The proposed framework assigns duty cycles to the sensor nodes based on their network traffic rate.
Lipismita Panigrahi, Raghab Ranjan Panigrahi, and Saroj Kumar Chandra
IEEE
Image captioning is implemented using Deep learning and NLP (Natural Language Processing) resulting in producing a description of an image. The proposed model generates a caption for an image using a Convolutional Neural Network (CNN) together with a Recurrent Neural Network (RNN) and area of attention. Previously, the image names were used as keys to map the images with descriptions. In order to achieve high performance, in the proposed model the image caption is based on the relationship between the areas of a picture (attention model), the words used in the caption, and the state of an RNN language model. The approach of progressive loading is employed for the loading of the image dataset. Further, for encoding the image dataset into a feature vector, VGG16 a pre-trained CNN is used. The extracted feature vector is given as input to the RNN model. These image encodings are output to a specific type of RNN model known as Long Short-Term Memory (LSTM) networks. Subsequently, the LSTM works on decoding the feature vector and predicts the sequence of words, resulting in the generation of descriptions or captions. The training performance is measured using one of the model’s quantitative analysis metrics known as BLEU.
Sourabh Kumar, Saroj Kumar Chandra, Ram Narayan Shukla, and Lipismita Panigrahi
IEEE
Automation has made tremendous changes in the industries. It has been used to automate the manual processes involved in different physical units of the industries. The purpose was to increase the production in the manufacturing. Now, Computers are being used in the industries to monitor functionalities of different production units with the help of artificial intelligence and internet of things (IoT). The IoT has revolutionized the industries. It is an interconnected network system of physical units. The core purpose of it to gather and share information among different physical units. The IoT has great impact on the many areas such as business, industry, medicine, the economy, transport, industrial robots and automation systems. IoT with artificial intelligence has wide range of industrial applications. Industry 4.0 is used in the industries where different industrial units are connected over the internet and interacting to make decisions via machine-to-machine communication. It has increased the benefits of industries in terms of production and supply chain management. Manufacturing industry monitors its production units in every 10 milliseconds to capture features of the product that is being produced. The features generated in this process are huge in amount. Critical observation is performed on the generated features to categorize the product as anomalous or good. Product classification is difficult task in the labeled datasets due to human bias in labeling the final product as anomalous or good. In this work, machine learning models is being used to detect and classify faulty product produced by manufacturing industry. Both qualitative and quantitative study will be carried out to compare various machine learning models.
Sunny Singh and Saroj Kumar Chandra
IEEE
Machine learning models have shown great performance in prediction and detection of many diseases such as cancer, heart attack, liver infection, and kidney infection. COVID-19 emerged as one of the deadly disease. Its cases grownin unpredictable manner. Regression is the mathematical technique in machine learning that can used to find relation between outcome variable with independent variable. In the present manuscript, regression has been used to predict COVID-19 growth. It has been found that the model is highly accurate in the COVID case prediction.
Uma Shankar Pandey, Gulshan Soni, and Saroj Kumar Chandra
IEEE
The modern healthcare monitoring system is getting benefited by new emerging technologies, such as tiny physiological sensors and the latest wireless technologies. The Wireless Body Area Networks (WBANs) introduce live tracing of vital signs, online diagnosis, and various other advantages when compared with the traditional system. The duty cycle (wake-up and sleep) performs a crucial function in the conservation of the limited battery power of physiological sensors. A scheme is introduced in this paper that adjusts the duration of superframe in the beaconenable mode of IEEE 802.15.4 MAC. The proposed scheme’s key objective is to prolong the lifetime of the physiological sensor by altering superframe duration. Through extensive simulations, it is shown that the superframe duration may be appropriately adjusted to improve energy efficiency without significantly reducing network throughput.
Sourabh Kumar, Saroj Kumar Chandra, Saurav Dixit, Kaushal Kumar, Shivam Kumar, Gunasekaran Murali, Nikolay Ivanovich Vatin, and Mohanad Muayad Sabri Sabri
MDPI AG
In the present study, Erosion wear of stainless steel 420 was predicted using an artificial neural network (ANN). Stainless steel 420 is used for making slurry transportation components, such as pump impellers and casings. The erosion wear performance was analyzed by using a slurry pot tester at the rotational speed of 600–1500 rpm with a time duration of 80–200 min. Fly ash was used as an erodent medium, and the solid concentration varied from 20 to 50%. The particle size of erodent selected for the erosion tests was <53 µm, 53–106 µm, 106–150 µm, 150–250 µm. A standard artificial neural network (ANN) for the prediction of erosion wear was designed using the MATLAB program. Erosion wear results obtained from experiments showed a good agreement with the ANN results. This technique helps in saving time and resources for a large number of experimental trials and successfully predicts the erosion wear rate of the coatings both within and beyond the experimental domain.
Saroj Kumar Chandra and Manish Kumar Bajpai
Springer Science and Business Media LLC
Saroj Kumar Chandra and Ashok Bhansali
IEEE
COVID-19 emerged as one of the major outbreak to human society and has been declared as pandemic by World Health Organization (WHO). The first phase was almost over in the month of February 2021 in India, but soon after second wave emerged out with greater impact. The whole world is struggling hard to contain the spread of virus but finding it difficult, as new mutations and variants are taking shape continuously. Many mathematical models have been designed to predict spread of COVID-19, but prediction fails due to evolution of virus as well as its behavior. WHO provided many guidelines to prevent spread of COVID-19 which includes social distancing, wearing of masks in public places and frequent sensitization of hands. Wearing of mask has been proved to be the most effective in preventing the spread of corona virus. Wearing masks perhaps is the most important life style change which could help contain the spread of virus specially in offices, malls, theaters, restaurants and other public places. Though the administration frequently issues guidelines to wear masks but it is really very difficult to identify the peoples without mask in large gatherings. In the present manuscript, an automatic mask detection system has been proposed using machine learning to automatically identity the people without masks.
Saroj Kumar Chandra, Abhishesk Shrivastava, and Manish Kumar Bajpai
Springer Singapore
Saroj Kumar Chandra, Avaneesh Singh, and Manish Kumar Bajpai
Springer Singapore
Avaneesh Singh, Saroj Kumar Chandra, and Manish Kumar Bajpai
Computers, Materials and Continua (Tech Science Press)
COVID-19 disease has emerged as one of the life threatening threat to the society A novel beta coronavirus causes it It began as unidentified pneumonia of unknown etiology in Wuhan City, Hubei province in China emerged in December 2019 No vaccine has been produced till now Mathematical models are used to study the impact of different measures used to decrease pandemic Mathematical models have been designed to estimate the numbers of spreaders in different scenarios in the present manuscript In the present manuscript, three different mathematical models have been proposed with different scenarios, such as screening, quarantine, and NPIs, to estimate the number of virus spreaders The analysis shows that the numbers of COVID-19 patients will be more without screening the peoples coming from other countries Since every people suffering from COVID-19 disease are spreaders The screening and quarantine with NPIs have been implemented to study their impact on the spreaders It has been found that NPI measures can reduce the number of spreaders The NPI measures reduce the spread function's growth and provide decision makers more time to prepare with in dealing with the disease © 2020 Tech Science Press All rights reserved
Saroj Kumar Chandra and Manish Kumar Bajpai
Elsevier BV
Saroj Kumar Chandra and Manish Kumar Bajpai
Springer Science and Business Media LLC
Saroj Kumar Chandra and Manish Kumar Bajpai
Elsevier BV
Muthineni Shireesha, Gargi Yadav, Saroj Kumar Chandra, and Manish Kumar Bajpai
IEEE
Image reconstruction with less projection data is one of the most challenging tasks in medical imaging. In the present manuscript, a novel image reconstruction model using Deep Convolutional Neural Network (CNN) is being presented where the whole data view is not available. The projection data has been used to restore missing information. This projection data is obtained by passing X-rays with different angles of rotation around the object to make an image. The excess of X-ray radiation in the human body can produce cancerous tissues. Hence, in the present work, fewer projection data has been used to reconstruct images without compromising the quality of the reconstructed image. Radon transform has been used to obtain projection data. CNN has been used in the present work to improve the quality, reduce artifacts and suppress noises in the reconstructed image.
Saroj Kumar Chandra and Manish Kumar Bajpai
Springer Science and Business Media LLC
Saroj Kumar Chandra and Manish Kumar Bajpai
Elsevier BV
Saroj Kumar Chandra and Manish Kumar Bajpai
IEEE
Benign brain tumor segmentation is one of the most challenging task. Benign tumors have very small differential characteristics to its surrounding non-tumorous cells. Variety of methods can be found in the literature for brain tumor segmentation, but most of the methods fails to detect such low differential data. In the present manuscript, fractional calculus based benign brain tumor segmentation method has been proposed. It has higher sensitivity towards low differential data. This feature has been exploited to segment benign brain tumor region. The performance of proposed method has been compared with existing state-of-the-art methods used for brain tumor segmentation. Higher performance has been achieved using proposed method in benign brain tumor segmentation.
Saroj Kumar Chandra and Manish Kumar Bajpai
IEEE
Brain tumor segmentation in early stage or benign stage is a complex and challenging task. It involves the process of identifying early stage brain tumor cells among normal healthy cells which have approximately similar characteristics or low variation to its surrounding tissues. Many methods have been proposed for benign brain tumor segmentation but most of the methods fails to detect such low variation data. In the present manuscript, fractional model for benign brain tumor segmentation is proposed. The model has higher sensitivity towards low differential data due to arbitrary order of derivative. This arbitrary order of derivative has been exploited to segment brain tumor in early stage. The results obtained have been compared with existing popular methods and it has been found that proposed method is superior among them in brain tumor segmentation in early stage.
Saroj Kumar Chandra and Manish Kumar Bajpai
IEEE
Image smoothing is one of the most useful application of image processing. The methods available for image smoothing can be divided into linear based (isotropic) and nonlinear based methods (anisotropic). Gaussian smoothing is one of the most linear model of image smoothing. It is unable to preserve edge details in the image. Partial differential equation(PDE) based methods are non-linear in nature and. They are able to preserve image boundaries. These methods have better trade-off between noise suppression and edge preserving. However, it is found that these methods creates side effects such as blocky effect, speckle effect after processing. The present manuscript presents a novel fractional calculus based approach for image smoothing. The model not only reduces the effect of noise but also preserves edges. The performance of proposed method has been validated by introducing various kinds of noises such as Gaussian, speckle and salt and pepper into image. Peak Signal to Noise Ratio (PSNR) and Structural Similarity Index Matrix (SSIM) have been used for quantitative evaluations. Higher values of PSNR and SSIM has been achieved by proposed method.