Padmaja B
@iare.ac.in
Associate Professor, Department of CSE
Institute of Aeronautical Engineering
Ms. B. Padmaja is a faculty member in the Department of Computer Science, Institute of Aeronautical Engineering, Hyderabad, Telangana, India. She has received her B.Tech from North Eastern Regional Institute of Science and Technology (NERIST), Arunachal Pradesh, India in 2001. She completed her M.Tech from School of IT, JNTUH, Hyderabad, India. Currently she is pursuing her research in “Reality Mining: Smart Phone Based Human Behavior Analysis” from JNTUH, Hyderabad. She is a member of ISTE and CSI. She has more than 18 years of teaching experience and published 20 research papers in various International Journals and Conferences.
EDUCATION
B.Tech (CSE): North Eastern Regional Institute of Science and Technology
M.Tech (CS): School of IT, JNTUH, Hyderabad
Ph.D (CSE): Pursuing from JNTUH, Hyderabad
RESEARCH INTERESTS
Machine Learning, Computer Vision, Social Network Analysis, Reality Mining
Scopus Publications
Scholar Citations
Scholar h-index
Scholar i10-index
Scopus Publications
Akella S. Narasimha Raju, K. Venkatesh, B. Padmaja, and G. Sucharitha Reddy
Springer Science and Business Media LLC
Bonthu Kotaiah, B. Padmaja, and Amitabha Yadav
AIP Publishing
B. Padmaja, B. J. D. Kalyani, Vindhya Chapala, Sharwan Solanki, and Abhiram Kaipa
AIP Publishing
Budi Padmaja, Myneni Madhu Bala, Epili Krishna Rao Patro, Adiraju Chaya Srikruthi, Vytla Avinash, and Chenumalla Sudheshna
Institute of Advanced Engineering and Science
Short message service (SMS) is one of the quickest and easiest ways used for communication, used by businesses, government organizations, and banks to send short messages to large groups of people. Categorization of SMS under different message types in their inboxes will provide a concise view for receivers. Former studies on the said problem are at the binary level as ham or spam which triggered the masking of specific messages that were useful to the end user but were treated as spam. Further, it is extended with multi labels such as ham, spam, and others which is not sufficient to meet all the necessities of end users. Hence, a multi-class SMS categorization is needed based on the semantics (information) embedded in it. This paper introduces an intelligent auto-response model using a semantic index for categorizing SMS messages into 5 categories: ham, spam, info, transactions, and one time password’s, using the multi-layer perceptron (MLP) algorithm. In this approach, each SMS is classified into one of the predefined categories. This experiment was conducted on the “multi-class SMS dataset” with 7,398 messages, which are differentiated into 5 classes. The accuracy obtained from the experiment was 97%.
Jayasri Kotti, B. Padmaja, and D. Deepa
River Publishers
Artificial Intelligence has become an essential part of modern technology. Although computer technology is advanced, it can still be improved to make it more user-friendly. One way to do this is to replace touchscreen desktops with a virtual mouse and keyboard. This can reduce the need for gadgets and enhance human-computer interaction. During the COVID-19 pandemic, reducing human intervention and dependency on devices has been critical in controlling the spread of the virus. This is where a battery-powered or Bluetooth mouse, powered by virtual reality technology, can be helpful. The virtual mouse is created using OpenCV and virtual reality technology, with the proposed system utilizing advanced tools such as MediaPipe and Python. The MediaPipe library is particularly useful in artificial intelligence projects, as it enhances the efficiency of the model. The system is an AI-based mouse and keyboard that can be controlled using hand gestures. The user interacts with the system through the camera output displayed on the screen, while the webcam serves as an input device. Python and OpenCV tools are used for implementation, making it applicable in pandemic situations and smart teaching systems. The proposed system works on Enhancing gesture Controlled Virtual Mouse and Virtual Keyboard through Virtual Assistant using AI Techniques.
B. Padmaja, CH. V. K. N. S. N. Moorthy, N. Venkateswarulu, and Myneni Madhu Bala
Springer Science and Business Media LLC
AbstractAutonomous cars have achieved exceptional growth in the automotive industry in the last century in terms of reliability, safety and affordability. Due to significant advancements in computing, communication and other technologies, today we are in the era of autonomous cars. A number of prototype models of autonomous cars have been tested covering several miles of test drives. Many prominent car manufacturers have started investing huge resources in this technology to make it commercialize in the near future years. But to achieve this goal still there are a number of technical and non-technical challenges that exist in terms of real-time implementation, consumer satisfaction, security and privacy concerns, policies and regulations. In summary, this survey paper presents a comprehensive and up-to-date overview of the latest developments in the field of autonomous cars, including cutting-edge technologies, innovative applications, and testing. It addresses the key obstacles and challenges hindering the progress of autonomous car development, making it a valuable resource for anyone interested in understanding the current state of the art and future potential of autonomous cars.
Haritha Akkineni, Madhu Bala Myneni, Budi Padmaja, Ananda Ravuri, CH. V. K. N. S. N. Moorthy, and Raviteja CMS
River Publishers
Privacy and anonymity aspects are playing a vital role in accessing smartphone apps. This is more evident in unexpected epidemic situations like COVID-19 while working with contact tracing apps. A human connectivity model is essential to analyse the widespread cases of viruses and vaccination patterns during the timeframe of March 2020 to May 2021. Smartphone apps that are supported by technologies like IoT and blockchain have already proven effective in tracing the Ebola epidemic. Thus, this technology, coupled with privacy-preserving features, would help to discover clusters with infectious contacts and alert the respective authorities. Besides, this can also allow us to understand the human connectivity model and the effectiveness of vaccines, which can aid in developing a plan of action for future epidemics. Hence, this article focuses on the analysis of data collected from contact tracing apps and a number of affected cases. It includes a study on early solutions with existing technologies, an overview and analysis of existing COVID-19 apps with vulnerabilities, proposed solutions, and data analysis on privacy and anonymity aspects of smartphone apps using the ARIMA model. It is evaluated by correlating it with the usage of contact tracing apps. The results assured a positive correlation between the number of downloads and the number of cases. This infers that even though the Indian government released these contact tracing apps, it all depends on the citizens to utilise them to their fullest. As a policy suggestion, it is stated that regardless of the prevalence of contact tracing apps, people must follow the rules and regulations suggested by the local health authorities and maintain social distancing in public places.
Padmaja B, Madhubala M, Nagaraju M, Nandhan Varma Somalaraju, Meghana Kovuri, and Krishnaveni Sriramwar
IEEE
Identification of respiratory disease is a vital step in respiratory disease diagnosis and treatment. Chest X-rays computed tomography (CT), and magnetic resonance imaging (MRI) scans are performed to evaluate the lungs and other constituents of the respiratory system. Chest X-rays are frequently used to assess respiratory diseases like pneumothorax, pneumonia, tuberculosis, and coronavirus disease. By examining the images, doctors can accurately identify the presence of certain conditions, such as abnormal cells, fluid build-up, and lung consolidation. Deep learning techniques can automatically analyze vast amounts of data and spot patterns that human experts might overlook. This may result in recommendations for treatments and diagnoses that are more precise. This proposed work aims to develop a reliable system that can classify chest X-ray images into Coronavirus Disease, Bacterial pneumonia, tuberculosis, and normal cases using convolutional neural networks (CNNs) which will be helpful in the medical field. We used a dataset that consists of 8000 images which belong to various classes. We trained our data on various pre-trained models like VGG-19, Inception Net V3, and ResNet 50 with various learning rates achieved an accuracy of 95%, 87%, and 98% respectively. This proposed work used Grad-CAM to provide insights into which areas the model paid more focus during classifications, which will assist health professionals in starting medication as immediately as possible.
B. Padmaja, G. Chandra Sekhar, Ch. V. Rama Padmaja, P. Chandana, and E. Krishna Rao Patro
Springer Nature Singapore
Suneetha Merugula, Buddi Padmaja, and Ragavi Veerubommu
Wiley
Heart disease detection through early‐stage syndrome remains as a main confront in present world situation. If it is not detected appropriate time, then this turns out to be the major cause of death. Several existing heart disease detection techniques are developed with lower detection performance and therefore it is very significant to introduce a novel heart disease detection model that poses the potential to detect heart disease from input data. A novel detection approach named, social water cycle algorithm‐based deep residual network (SWCA‐based DRN) is proposed for classification of heart disease. The developed SWCA algorithm is a newly designed by the hybridization of social optimization algorithm and water cycle algorithm. Here, an input data is initially preprocessed and the feature fusion procedure is carried out RV coefficient enabled rider optimization algorithm‐based neural network. With the fused feature result, heart disease classification is performed utilizing a DRN classifier where training procedure of DRN is done by proposed optimization algorithm, named SWCA. Furthermore, developed SWCA‐enabled DRN technique outperformed different other present heart disease detection approaches and attained superior performance concerning the performance measures, like testing accuracy, sensitivity, and specificity with highest values of 0.941, 0.954, and 0.925.
B Padmaja, K Sai Sravan, E Krishna Rao Patro, and G Chandra Sekhar
IOP Publishing
Abstract Cyber security is the major concern in today’s world. Over the past couple of decades, the internet has grown to such an extent that almost every individual living on this planet has the access to the internet today. This can be viewed as one of the major achievements in the human race, but on the flip side of the coin, this gave rise to a lot of security issues for every individual or the company that is accessing the web through the internet. Hackers have become active and are always monitoring the networks to grab every possible opportunity to attack a system and make the best fortune out of its vulnerabilities. To safeguard people’s and organization’s privacy in this cyberspace, different network intrusion detection systems have been developed to detect the hacker’s presence in the networks. These systems fall under signature based and anomaly based intrusion detection systems. This paper deals with using anomaly based intrusion detection technique to develop an automation system to both train and test supervised machine learning models, which is developed to classify real time network traffic as to whether it is malicious or not. Currently the best models by considering both detection success rate and the false positives rate are Artificial Neural Networks(ANN) followed by Support Vector Machines(SVM). In this paper, it is verified that Artificial Neural Network (ANN) based machine learning with wrapper feature selection outperforms support vector machine (SVM) technique while classifying network traffic as harmful or harmless. Initially to evaluate the performance of the system, NSL-KDD dataset is used to train and test the SVM and ANN models and finally classify real time network traffic using these models. This system can be used to carry out model building automatically on the new datasets and also for classifying the behaviour of the provided dataset without having to code.
B Padmaja, P Naga Shyam Bhargav, H Ganga Sagar, B Diwakar Nayak, and M Bhushan Rao
IOP Publishing
Abstract Visually impaired and senior citizens find it difficult to identify different banknotes, driving the need for an automated system to recognize currency notes. This study proposes recognizing Indian currency notes of various denominations using Deep Learning through the CNN model. While not recognizing currency notes is one issue, identifying fake notes is another major issue. Currency counterfeiting is the illegal imitation of currency to deceive its recipient. The current existing methodologies for identifying a phony note rely on hardware. A method completely devoid of hardware that relies on specific security features to help distinguish a legitimate currency note from an illegitimate one is much needed. These features are extracted using the boundary box region of interest (ROI) and Canny Edge detection in OpenCV implemented in Python, and the multi scale template matching algorithm is applied to match the security features and differentiate fake notes from legitimate notes.
Budi Padmaja, Madhu Bala Myneni, and Epili Krishna Rao Patro
Springer Science and Business Media LLC
Multi activity-multi object recognition (MAMO) is a challenging task in visual systems for monitoring, recognizing and alerting in various public places, such as universities, hospitals and airports. While both academic and commercial researchers are aiming towards automatic tracking of human activities in intelligent video surveillance using deep learning frameworks. This is required for many real time applications to detect unusual/suspicious activities like tracking of suspicious behaviour in crime events etc. The primary purpose of this paper is to render a multi class activity prediction in individuals as well as groups from video sequences by using the state-of-the-art object detector You Look only Once (YOLOv3). By optimum utilization of the geographical information of cameras and YOLO object detection framework, a Deep Landmark model recognize a simple to complex human actions on gray scale to RGB image frames of video sequences. This model is tested and compared with various benchmark datasets and found to be the most precise model for detecting human activities in video streams. Upon analysing the experimental results, it has been observed that the proposed method shows superior performance as well as high accuracy.
Gangannagari Upender Reddy, Myneni Madhu Bala, and B Padmaja
IEEE
Image Forensics has lot of importance in digital forensics in crime investigation process. As a coin has two different sides, many anti-Forensic tools are available to help criminals for hiding traces of forgery which also have been evolved with advancement in modern technology. In this paper the major focus is on overview of available forensics tools and frequent image processing techniques involved in it to investigate crime related digital traces.
Adusumilli Taraka Venkata Satya Sai Rohit, B. Padmaja, Y. Mohana Roopa, and E. Krishna Rao Patro
Springer Singapore
Sai Charan Pedduri, Sai Ruchith Reddy Ginnavaram, Madhu Bala Myneni, and B. Padmaja
Springer Singapore
B Padmaja, V Prasa, and K Sunitha
European Alliance for Innovation n.o.
INTRODUCTION: Automatic detection and recognition of various human physical movements while performing daily life activities such as walking, jogging, running, sitting, standing etc. are usually considered as Activity Recognition (AR). AR is a prominent research area in many applications, such as elderly care, security and surveillance, smart homes, health and fitness. Extremely Randomized Trees Classifier (ET Classifier) is a type of ensemble learning technique used in Activity Recognition, which clusters several different decision trees into a forest from a single learning set and gives the classification result. But it suffers from high variance and over-fitting problem due to high inter-dependency among hyperparameters during model building. OBJECTIVES: The primary objective of this paper is to propose a novel Random_Split_Point procedure for Extra tree classifier to make the existing approach more robust, less variance, less computational time in obtaining optimal split points and faster in model building. This approach generates K random split points from all the candidate features of the dataset and selects the best split point based on the maximum score obtained by information gain measure. METHODS: In the proposed method to improve the randomization and accuracy of AR system, a novel random split-point procedure for ET classifier is proposed. This approach reduces the bias-variance problem induced due to the three hyperparameters such as K, nmin and M used in split-point procedure of existing ET classifier (K : number of randomly selected attributes at each node, nmin : minimum sample size for splitting a node, M : number of decision trees for ensemble). This approach generates K random split points from all the candidate features of the dataset and selects the best split point based on the maximum score obtained by information gain measure. RESULTS: The proposed approach is experimented with two public AR datasets HAR and HAPT (UCI Machine Learning Repository) containing 6 and 12 activities respectively. In HAR dataset, smartphone sensed sensor signals of 3 static and 3 dynamic human daily activities are there, where as in HAPT dataset apart from these 6 daily activities, 6 postural transitions data is available. Experimental results and comparative analysis show that the proposed method outperforms over other existing techniques with an accuracy of 94.16% for HAR dataset and 92.63% for HAPT dataset. It also takes less computational time in finding optimal split-points and less model building time. CONCLUSION: AR systems can be used as an intelligent system in healthcare to monitor the behaviour of healthy people by recognizing their daily activities. These systems also help in early detection of some chronic diseases and improve the quality of life. In this paper, an attempt is made to improve the accuracy of Activity Recognition over some existing methods.
P. Sri Harsha Vardhan Goud, Y. Mohana Roopa, and B. Padmaja
IEEE
Sports are the most important recreational activity. Sports are of many types. Some may be played individually, while some are played in teams. Every country wants to get fame at the global level in different sports. In order to achieve fame, countries are investing in sports and games to enhance the performance of their teams and players. Many people are involved in the analysis of the performances in a sport like notational analyst, who make strategies and tactics for a game; bio-mechanist, who takes the responsibility of fitness of players and tries to get extraordinary results; team managers and coaches. With the advent of machine learning in sports, there is a lot of improvement in the analysis of performances. In mere future, the teams may not have coaches to analyze their performances. In this paper, I am going to discuss about the analysis role of machine learning in the improvement of performances of players and the team in different sports and how the wearable technology helps the players to know their performance levels and further improvements
R. Shreyas, B. Padmaja, H. B. Adithya, and M. P. Sunil
Springer International Publishing
Agriculture has been evolving over the past century. We can see a rise in the agricultural yield over the years. But, the existing techniques are turning out to be less effective, especially with the rapidly increasing population. The modern agricultural methods will need a radical transformation if they’re going to keep up. Hence, there is a need to automate agriculture. We propose a system that a farmer can make use of, to help him in agricultural applications. An autonomous ground vehicle (AGV) is designed to monitor the presence of moisture in the soil, detect and control pest, and for perimeter surveillance. Also, there is an option to live-stream the videos on their phone, tab or laptop from the onboard camera on the AGV.
Budi Padmaja, Venkata Rama Prasad Vaddella, and Kota Venkata Naga Sunitha
Science Publications
Sensor-based human motion detection requires the subtle amount of knowledge about various human activities from fitted sensor observations and readings. The prevalent pattern recognition methodologies have made immense progress over recent years. Nonetheless, these kind of methods usually rely on the particular heuristic variable extraction, which could inhibit generalization realization. This paper presents a distributed and parallel decision forest approach for modeling the Human Activity Recognition Using Smartphones Data. We made an attempt to achieve an optimal generalization performance with possible reduction in overfitting. Later, we compared the performance of proposed procedure with some existing approaches. It is observed that our adopted procedure outperforms with comparatively better statistical performance measures. It also gained 4.7x speed up in computation.
B. Padmaja, V. V. Rama Prasad, K. V. N. Sunitha, N. Chandra Sekhar Reddy, and C. H. Anil
Springer Singapore
Stress has become an inevitable part of human lives and a major concern for public health. Especially, in today’s highly competitive world, stress levels of individuals have increased and have pervaded their work life. In the workplace, a professional has to relentlessly confront a plethora of situations and issues such as work pressure, deadlines, disaster management, adapting to new changes. The incessant stress can cause many health issues, such as high blood pressure, insomnia, vulnerability to infections, and heart diseases. Our aim is to design a cognitive stress-level detection system (DetectStress) which unobtrusively assess an individual’s stress levels based on smartphone daily activity data and wireless physical activity tracker (FITBIT) data. The device FITBIT records an individual’s daily logs of food, weight, sleep patterns, heart rate, and physical activities. Individual stress was also measured using the most preferred psychological instrument perceived stress scale (PSS) questionnaire on monthly basis. The data was gathered using an online form which consists of ten questions in several categories, and these questions ask about their feelings and thoughts during the last month. For this study, data was collected from 35 young adults for a period of 2 months. The data includes their social behavior and other routine activities collected from smartphone. Our system uses machine learning approach for stress detection along with perceived stress scale questionnaire score (PSS). The model is evaluated using two classifiers such as Naive Bayes and Decision Tree and compared against a baseline classifier random classifier. Naive Bayes classifier has increased in performance from 55 to 72% in terms of accuracy than other models in detecting stress levels (low < med < high). This paper gives the scope for stress detection more accurately using smartphone sensor technology rather than clinical experimentations. This system is assessed in real time with young college students in India. This paper also proposes the architecture of stress based on both physiological and behavioral response of individuals. The uniqueness of this work lies in the simplification of the stress detection process.
B Padmaja, P V Narasimha Rao, M Madhu Bala, and E Krishna Rao Patro
IEEE
Autonomous car is a ground vehicle that is capable of driving without user interference. Traffic congestion and number of collisions are major issues in road traffic control due to rapid increase day-by-day. Autonomous cars provide a solution to this problem in an efficient and economical way. The proposed system utilizes mathematical models like neural networks and image processing techniques to sense the environment. This is implemented as three major components: curved road detection (steering), road sign and signal detection and obstacle detection (collision avoidance). Back Propagation is used for steering control with detection of curved roads; Haar features are used for road signal, sign detection and a distance sensor for collision avoidance. Data collected from the sensors is sent to a server for processing. Based on the result, a command is sent to the car. A GPS module attached to the car identifies the location of the car and with the help of a 3rd party location service, route to destination is identified and directions are sent to the car. Wireless networks are used to transmit data between sensors and the server. Python scripts are used to control and integrate all the units together. The designed system can attain high accuracy with real – time constraints.