Browsing by Author "Manjari Gupta"

Now showing 1 - 20 of 56

A Comparative Performance Analysis of Activation Functions for Cardiovascular Disease Detection Using ECG Images
(John Wiley and Sons Inc, 2025) Mrityunjay Chaubey; Abhay Kumar Pathak; Marisha; Manjari Gupta
In recent years, artificial intelligence (AI) has become an automated tool for detecting cardiovascular diseases using ECG images. Activation functions are the core of neural network models, ranging from shallow to deep convolutional neural networks (CNN). In ECG image-based cardiovascular disease detection, activation functions enable the network to capture non-linear patterns like irregular heartbeats and subtle anomalies. The proposed CNN architecture in this paper comprised convolutional layers for feature extraction, followed by custom activation functions to introduce non-linearity and enhanced learning. These features are downsampled using max pooling and aggregated through global average pooling. Fully connected layers, with a suitable dropout regularization, map the features to the final classification output, which is probabilistically determined using a softmax activation function. This paper used a public dataset of ECG images of cardiac patients to analyze the significance of activation functions in predicting the four main cardiac abnormalities: irregular heartbeat, myocardial infarction, history of myocardial infarction, and normal person classes. We have analyzed 19 different activation functions for their detection performance on the same dataset. The detection performance is compared with the existing state-of-the-art studies. A set of activation functions is suggested for robust and accurate detection of cardiovascular disease using ECG images. © 2025 Wiley-VCH GmbH.
A Decision Tree Approach for Design Patterns Detection by Subgraph Isomorphism
(2010) Akshara Pande; Manjari Gupta; A.K. Tripathi
In many object oriented softwares, there are recurring patterns of classes. Design pattern instances are important for program understanding and software maintenance.Hence a reliable design pattern mining is required. Here we are applying decision tree approach followed by subgraph isomorphism technique for design pattern detection. © Springer-Verlag Berlin Heidelberg 2010.
A Deep Learning Based Model to Study the Influence of Different Brain Wave Frequencies for the Disorder of Depression
(Springer Science and Business Media Deutschland GmbH, 2023) Bethany Gosala; Emmanuel Raj Gosala; Manjari Gupta
The human brain is one of the most advanced, complex, and incredible machines which has continued to fascinate scientists, researchers, and scholars for hundreds of years. Many experiments and studies have been done on the human brain to understand its mechanism and how it works, yet we are not close to understanding its full potential. One way of studying the brain is to study the brain wave frequencies which are emitted by it. The brain emits five different types of waves namely, delta, theta, alpha, beta, and gamma. Studying these different waves can help in solving various psychological issues, and problems like anxiety, stress, and depression which every human faces at least once in their life, according to WHO depression will be the main cause of mental illness by 2030. This work aims to find the influence of different brain waves and their involvement in the case of depression. For this, we have used deep learning techniques and developed a supervised learning model called convolutional neural network (CNN) for the classification of signals from Major Depression Disorder (MDD) from the healthy control. The developed CNN is run in five brain waves, and we calculated the accuracy for performance evaluation of the developed model for each brain wave frequency. The best accuracy we get is 98.4% for the delta wave followed by 97.6% for the alpha wave and the beta wave giving the least accuracy of 72.83%. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.
A new approach for detecting design patterns by graph decomposition and graph isomorphism
(2010) Akshara Pande; Manjari Gupta; A.K. Tripathi
Design Pattern Detection is a part of many solutions to Software Engineering difficulties. It is a part of reengineering process and thus gives important information to the designer. Design Pattern existence improve the program understanding and software maintenance. With the help of these patterns specific design problem can be solved and object oriented design become more flexible and reusable. Hence a reliable design pattern mining is required. Here we are applying graph decomposition followed by graph isomorphism technique for design pattern detection. © 2010 Springer-Verlag Berlin Heidelberg.
A novel approach for code smell detection: An empirical study
(Institute of Electrical and Electronics Engineers Inc., 2021) Seema Dewangan; Rajwant Singh Rao; Alok Mishra; Manjari Gupta
Code smells detection helps in improving understandability and maintainability of software while reducing the chances of system failure.In this study, six machine learning algorithms have been applied to predict code smells.For this purpose, four code smell datasets (God-class, Data-class, Feature-envy, and Long-method) are considered which are generated from 74 open-source systems.To evaluate the performance of machine learning algorithms on these code smell datasets, 10-fold cross validation technique is applied that predicts the model by partitioning the original dataset into a training set to train the model and test set to evaluate it.Two feature selection techniques are applied to enhance our prediction accuracy.The Chi-squared and Wrapper-based feature selection techniques are used to improve the accuracy of total six machine learning methods by choosing the top metrics in each dataset.Results obtained by applying these two feature selection techniques are compared.To improve the accuracy of these algorithms, grid search-based parameter optimization technique is applied.In this study, 100% accuracy was obtained for the Long-method dataset by using the Logistic Regression algorithm with all features while the worst performance 95.20% was obtained by Naive Bayes algorithm for the Long-method dataset using the chi-square feature selection technique. © 2021 Institute of Electrical and Electronics Engineers Inc.. All rights reserved.
A Reliable People Tracking in Nuclear Power Plant Control Room Monitoring System Using Particle Filter
(Institute of Electrical and Electronics Engineers Inc., 2025) Mrityunjay Chaubey; Lalit Kumar Singh; Manjari Gupta; Pooja Singh
The control room functions as the core nervous system of a nuclear power plant (NPP), emphasizing the crucial need for real-time monitoring of all activities inside to guarantee comprehensive safety. The maintenance of a high level of reliability in the real-time monitoring system within the control room of an NPP is of utmost importance in order to effectively mitigate any potential failures that may occur during the monitoring process. The software and hardware problems can both cause unplanned outages in a large-scale distributed monitoring system. To address the challenge of NPP control room monitoring, a particle filtering-based people tracking system for NPP control room monitoring is introduced to ensure the safety, security, and reliability of the NPP control room. In addition to tracking people for the monitoring of the NPP control room, the suggested technique also provides a reliability study of the large-distributed monitoring system. © 1963-2012 IEEE.
A Reliable People Tracking in Nuclear Power Plant Control Room Monitoring System Using Particle Filter
(Institute of Electrical and Electronics Engineers Inc., 2024) Mrityunjay Chaubey; Lalit Kumar Singh; Manjari Gupta; Pooja Singh
The control room functions as the core nervous system of a nuclear power plant (NPP), emphasizing the crucial need for real-time monitoring of all activities inside to guarantee comprehensive safety. The maintenance of a high level of reliability in the real-time monitoring system within the control room of an NPP is of utmost importance in order to effectively mitigate any potential failures that may occur during the monitoring process. The software and hardware problems can both cause unplanned outages in a large-scale distributed monitoring system. To address the challenge of NPP control room monitoring, a particle filtering-based people tracking system for NPP control room monitoring is introduced to ensure the safety, security, and reliability of the NPP control room. In addition to tracking people for the monitoring of the NPP control room, the suggested technique also provides a reliability study of the large-distributed monitoring system. IEEE
A Review of Deep Learning-based Human Activity Recognition on Benchmark Video Datasets
(Taylor and Francis Ltd., 2022) Vijeta Sharma; Manjari Gupta; Anil Kumar Pandey; Deepti Mishra; Ajai Kumar
Different types of research have been done on video data using Artificial Intelligence (AI) deep learning techniques. Most of them are behavior analysis, scene understanding, scene labeling, human activity recognition (HAR), object localization, and event recognition. Among all these, HAR is one of the challenging tasks and thrust areas of video data processing research. HAR is applicable in different areas, such as video surveillance systems, human-computer interaction, human behavior characterization, and robotics. This paper aims to present a comparative review of vision-based human activity recognition with the main focus on deep learning techniques on various benchmark video datasets comprehensively. We propose a new taxonomy for categorizing the literature as CNN and RNN-based approaches. We further divide these approaches into four sub-categories and present various methodologies with their experimental datasets and efficiency. A short comparison is also made with the handcrafted feature-based approach and its fusion with deep learning to show the evolution of HAR methods. Finally, we discuss future research directions and some open challenges on human activity recognition. The objective of this survey is to give the current progress of vision-based deep learning HAR methods with the up-to-date study of literature. © 2022 The Author(s). Published with license by Taylor & Francis Group, LLC.
A review of missing video frame estimation techniques for their suitability analysis in NPP
(Korean Nuclear Society, 2022) Mrityunjay Chaubey; Lalit Kumar Singh; Manjari Gupta
The application of video processing techniques are useful for the safety of nuclear power plants by tracking the people online on video to estimate the dose received by staff during work in nuclear plants. Nuclear reactors remotely visually controlled to evaluate the plant's condition using video processing techniques. Internal reactor components should be frequently inspected but in current scenario however involves human technicians, who review inspection videos and identify the costly, time-consuming and subjective cracks on metallic surfaces of underwater components. In case, if any frame of the inspection video degraded/corrupted/missed due to noise or any other factor, then it may cause serious safety issue. The problem of missing/degraded/corrupted video frame estimation is a challenging problem till date. In this paper a systematic literature review on video processing techniques is carried out, to perform their suitability analysis for NPP applications. The limitation of existing approaches are also identified along with a roadmap to overcome these limitations. © 2021
A study for method-level code smells detection using machine learning algorithms
(Academic Press, 2025) Rajwant Singh Rao; Seema Dewangan; Alok Mishra; Manjari Gupta
Motivation: Code smells reflect poor design decisions that degrade software quality and maintainability. Although several machine learning algorithms have been proposed to detect code smells, the impact of feature selection and cross-validation on certain method-level smells, specifically Long Parameter List and Switch Statements, has not been adequately explored in prior research. Methodology: This study employs a rigorous methodology to investigate the detection of four method-level code smells—Long Parameter List (LPL), Switch Statement (SS), Feature Envy (FE), and Long Method (LM) using twenty machine learning algorithms. We apply the Information Gain feature selection algorithm and the Equal Width Discretization (EWD) class balancing method. Performance is evaluated using 10-fold cross-validation across multiple metrics: accuracy, precision, recall, F-measure, MCC, ROC-area, and PRC-area. Key Findings: The proposed framework achieved a remarkable 99.77% accuracy for the Long Method dataset using the Filtered Classifier with feature selection and class balancing. Importantly, this study is the first to demonstrate the effect of feature selection and cross-validation on the LPL and SS datasets, where significant performance improvements are also observed. Contributions: A comprehensive comparative analysis of 20 machine learning algorithms on four method-level code smell datasets. © 2025 The Author(s)
A study of dealing class imbalance problem with machine learning methods for code smell severity detection using PCA-based feature selection technique
(Nature Research, 2023) Rajwant Singh Rao; Seema Dewangan; Alok Mishra; Manjari Gupta
Detecting code smells may be highly helpful for reducing maintenance costs and raising source code quality. Code smells facilitate developers or researchers to understand several types of design flaws. Code smells with high severity can cause significant problems for the software and may cause challenges for the system's maintainability. It is quite essential to assess the severity of the code smells detected in software, as it prioritizes refactoring efforts. The class imbalance problem also further enhances the difficulties in code smell severity detection. In this study, four code smell severity datasets (Data class, God class, Feature envy, and Long method) are selected to detect code smell severity. In this work, an effort is made to address the issue of class imbalance, for which, the Synthetic Minority Oversampling Technique (SMOTE) class balancing technique is applied. Each dataset's relevant features are chosen using a feature selection technique based on principal component analysis. The severity of code smells is determined using five machine learning techniques: K-nearest neighbor, Random forest, Decision tree, Multi-layer Perceptron, and Logistic Regression. This study obtained the 0.99 severity accuracy score with the Random forest and Decision tree approach with the Long method code smell. The model's performance is compared based on its accuracy and three other performance measurements (Precision, Recall, and F-measure) to estimate severity classification models. The impact of performance is also compared and presented with and without applying SMOTE. The results obtained in the study are promising and can be beneficial for paving the way for further studies in this area. © 2023, Springer Nature Limited.
A Transformer Based Emotion Recognition Model for Social Robots Using Topographical Maps Generated from EEG Signals
(Springer Science and Business Media Deutschland GmbH, 2024) Gosala Bethany; Manjari Gupta
Emotions are an integral part of living beings which influence their thoughts, actions, and interactions with other beings. Understanding human emotions is very important in communicating with others. Developing an emotion recognition model that can be implemented in robots is a critical step in human-robot interaction (HRI). With the rise of artificial intelligence, many techniques are available in machine learning and deep learning to solve this problem, one such technique is Transformers. Transformers, are used in trending technologies like BERT, ChatGPT, DALL-E-2, etc., We used transformers in this study as they have an edge over other by providing flexibility, adaptability, transfer learning, multimodality, parallelization etc., The dataset used is GAMEEMO, which contains EEG signals which are collected from 28 subjects while they were playing four computer-based games which emulate emotions like boring, calm, horror, and funny. Using EEG signal for emotion recognition have advantages like direct measure of brain activity, non-invasiveness, good temporal resolution etc., First, we preprocessed the raw EEG signal using bandpass filtering then created a 5-s epoch out of signal. Next, we converted the 1D EEG signal to a 2D topographical image using independent component analysis by taking 10 principal components out of 14 by persevering at least 95% of the variance in the data. From 9 h and 20 min of GAMEEMO EEG signal, we generated 82, 880 topographical images. Finally, these images were fed to a deep learning-based visual transformers model for the classification of emotions, the best accuracy of the model is 84.71%, our model performed better when compared with the other state-of-the-art models. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2024.
Advanced Brain Tumor Classification from MRI Images with Vision and Swin Transformer Models
(Institute of Electrical and Electronics Engineers Inc., 2025) Anjali Jain; Ankita Mishra; Bethany Gosala; Manjari Gupta; Alwin Poulose
Brain tumors, characterized by abnormal cell growth within the brain, represent one of the most life-threatening conditions due to their potential to severely impair brain function and cause neurological complications, including death. Accurate classification of brain tumors is crucial in medical diagnosis, as misdiagnosis can lead to ineffective treatment and reduced patient survival. Magnetic Resonance Imaging (MRI) is a widely utilized non-invasive technique for acquiring high-contrast grayscale brain images commonly employed in tumor diagnosis. This study uses MRI images to evaluate the effectiveness of two pre-trained transformer-based models for brain tumour classification, specifically the Vision Transformer and Swin Transformer. The models were trained and tested on a publicly available Kaggle dataset containing 3,000 brain MRI scans. Our methodology employed the fine-tuned Vision transformer and Swin Transformer on the MRI dataset for classification. The models' performance was assessed using accuracy, precision, recall, and F1-score metrics. The Vision Transformer achieved an accuracy of 95.30%, while the Swin Transformer outperformed it with an accuracy of 98.71%. These results highlight the efficacy of transformer models in accurately identifying common brain tumors, demonstrating their potential to enhance computer-aided diagnosis and support healthcare professionals in making swift and accurate diagnostic decisions. © 2025 IEEE.
Analysis of qRT-PCR Data to Identify the Most Stable Reference Gene Using gQuant
(Bio-protocol LLC, 2025) Abhay Kumar Pathak; Sukhad Kural; Shweta K. Singh; Lalit Kumar; Manjari Gupta; Garima Jain
The accurate quantification of nucleic acid–based biomarkers, including long non-coding RNAs (lncRNAs), messenger RNAs (mRNAs), and microRNAs (miRNAs), is essential for disease diagnostics and risk assessment across the biological spectrum. Quantitative reverse transcription PCR (qRT-PCR) is the gold standard assay for the quantitative measurement of RNA expression levels, but its reliability depends on selecting stable reference targets for normalization. Yet, the lack of consensus on a universally accepted reference gene for a given sample type or species, despite being necessary for accurate quantification, presents a challenge to the broad application of such biomarkers. Various tools are currently being used to identify a stably expressed gene by using qRT-PCR data of a few potential normalizer genes. However, existing tools for normalizer gene selection are fraught with both statistical limitations and inadequate graphical user interfaces for data visualization. gQuant, the tool presented here, essentially overcomes these limitations. The tool is structured in two key components: the preprocessing component and the data analysis component. The preprocessing addresses missing values in the given dataset by the imputation strategies. After data preprocessing, normalizer genes are ranked using democratic strategies that integrate predictions from multiple statistical methods. The effectiveness of gQuant was validated through data available online as well as in-house data derived from urinary exosomal miRNA expression datasets. Comparative analysis against existing tools demonstrated that gQuant delivers more stable and consistent rankings of normalizer genes. With its promising performance, gQuant enhances the precision and reproducibility in the identification of normalizer genes across diverse research scenarios, addressing key limitations of RNA biomarker–based translational research. © 2025 The Authors.
Automatic classification of uml class diagrams using deep learning technique: Convolutional neural network
(MDPI AG, 2021) Bethany Gosala; Sripriya Roy Chowdhuri; Jyoti Singh; Manjari Gupta; Alok Mishra
Unified Modeling Language (UML) includes various types of diagrams that help to study, analyze, document, design, or develop any software efficiently. Therefore, UML diagrams are of great advantage for researchers, software developers, and academicians. Class diagrams are the most widely used UML diagrams for this purpose. Despite its recognition as a standard modeling language for Object-Oriented software, it is difficult to learn. Although there exist repositories that aids the users with the collection of UML diagrams, there is still much more to explore and develop in this domain. The objective of our research was to develop a tool that can automatically classify the images as UML class diagrams and non-UML class diagrams. Earlier research used Machine Learning techniques for classifying class diagrams. Thus, they are required to identify image features and investigate the impact of these features on the UML class diagrams classification problem. We developed a new approach for automatically classifying class diagrams using the approach of Convolutional Neural Network under the domain of Deep Learning. We have applied the code on Convolutional Neural Networks with and without the Regularization technique. Our tool receives JPEG/PNG/GIF/TIFF images as input and predicts whether it is a UML class diagram image or not. There is no need to tag images of class diagrams as UML class diagrams in our dataset. © 2021 by the authors. Licensee MDPI, Basel, Switzerland.
Code Smell Detection Using Ensemble Machine Learning Algorithms
(MDPI, 2022) Seema Dewangan; Rajwant Singh Rao; Alok Mishra; Manjari Gupta
Code smells are the result of not following software engineering principles during software development, especially in the design and coding phase. It leads to low maintainability. To evaluate the quality of software and its maintainability, code smell detection can be helpful. Many machine learning algorithms are being used to detect code smells. In this study, we applied five ensemble machine learning and two deep learning algorithms to detect code smells. Four code smell datasets were analyzed: the Data class, the God class, the Feature-envy, and the Long-method datasets. In previous works, machine learning and stacking ensemble learning algorithms were applied to this dataset and the results found were acceptable, but there is scope of improvement. A class balancing technique (SMOTE) was applied to handle the class imbalance problem in the datasets. The Chi-square feature extraction technique was applied to select the more relevant features in each dataset. All five algorithms obtained the highest accuracy—100% for the Long-method dataset with the different selected sets of metrics, and the poorest accuracy, 91.45%, was achieved by the Max voting method for the Feature-envy dataset for the selected twelve sets of metrics. © 2022 by the authors.
Design Pattern Detection by normalized cross correlation
(2010) Manjari Gupta; Akshara Pande; Rajwant Singh Rao; A.K. Tripathi
Design Pattern Detection has been documented so far in the literature. The knowledge of design Pattern existence in the program improves the program understanding and software maintenance. Design pattern is a technology for design reuse. Experts store their experiences in the form of design patterns. Reengineering done by novice users will be successful if a reliable design pattern mining exists. There are 23 design patterns defined by experts. Here we are taking the UML diagrams corresponding to design pattern and corresponding to source code. Our main aim is to find out whether a particular design pattern exists in system design (source code) or not. For this we have extracted the relationship graphs (consisting of nodes and edges), and then tried to detect the design pattern. In this paper we have applied normalized cross correlation and taking design pattern as a template tried to find out its existence in system design. Normalized cross correlation (NCC) has been used extensively for many machine vision applications. Normalized cross correlation has been commonly used to evaluate the degree of similarity or dissimilarity between two images. © 2010 IEEE.
Design pattern detection using dpdetect algorithm
(Blue Eyes Intelligence Engineering and Sciences Publication, 2019) Jyoti Singh; Manjari Gupta
Design Patterns help to solve several recurring design issues in object oriented software. Expert software designers give design in terms of already proven design patterns to make their design more standard and less error-prone. Having the knowledge of design patterns used in the design helps to get insight into design. Thus detection of Design Patterns is very important for software designer to get significant information during re-engineering process. Detecting Design Patterns from source code or design of software system will help in the software understanding and its maintenance. It is also useful for novice developers who can get the idea about how to give solution (design) of any particular application using design patterns proposed by expert designers. Many design patterns detection approaches have been proposed by different researchers working in this field for more than two decades. These approaches consider structural, behavioural and/or semantic analysis of software system. Many sub graph isomorphism techniques were used to detect design patterns in case of structural analysis. In this paper we are using a branch and bound with backtracking algorithm for sub graph isomorphism, proposed by Asiler and Yazici [19]. We use this algorithm to show how this recover all the instances of design patterns from system design(renamed as DPDetect). Our main aim is to detect whether a particular design instance of design pattern is found in system design or not. It uses structural aspects of design patterns so it is based on only static analysis. © BEIESP.
Design Pattern Detection using inexact graph matching
(2010) Manjari Gupta; Rajwant Singh Rao; Anil Kumar Tripathi
Design Patterns are proven solution to common recurring design problems. Design Pattern Detection is most important activity that may support a lot to re-engineering process and thus gives significant information to the designer. Knowledge of design pattern exists in the system design improves the program understanding and software maintenance. Therefore, an automatic and reliable design pattern discovery is required. Graph theoretic approaches have been used for design pattern detection in past. Here we are applying an algorithm which decomposes the graph matching process into K phases, where the value of K ranges from 1 to the minimum of the numbers of nodes in the two graphs to be matched. The effectiveness of this algorithm results from the use of small values of K, and significantly reduces the search and space and producing very good matching between graphs. The same algorithm we are here using for design pattern detection from the system design. © 2010 Kongu Engineering College.
Design pattern mining for GIS application using graph matching techniques
(2010) Akshara Pande; Manjari Gupta; A.K. Tripathi
Design Pattern Detection is a part of many solutions to Software Engineering difficulties. It is a part of reengineering process and thus gives important information to the designer. Design Pattern existence improve the program understanding and software maintenance. With the help of these patterns specific design problem can be solved and object oriented design become more flexible and reusable. Hence a reliable design pattern mining is required. A GIS is an information system designed to work with data referenced by spatial / geographical coordinates. Here we are detecting design patterns so that it can be used as a conceptual tool to cope with recurrent problems appearing in the GIS domain. In this way, GIS applications can evolve smoothly, because maintenance is achieved by focusing on different concerns at different times. © 2010 IEEE.