Bhagwati Sharan
@srmap.edu.in
Department of Computer Science and engineering
SRM University-AP
RESEARCH, TEACHING, or OTHER INTERESTS
Computer Engineering, Computer Networks and Communications, Multidisciplinary
Scopus Publications
Scholar Citations
Scholar h-index
Scopus Publications
Megha Chhabra, Bhagwati Sharan, May Elbarachi, and Manoj Kumar
Springer Science and Business Media LLC
AbstractThis study aims to improve the performance of organic to recyclable waste through deep learning techniques. Negative impacts on environmental and Social development have been observed relating to the poor waste segregation schemes. Separating organic waste from recyclable waste can lead to a faster and more effective recycling process. Manual waste classification is a time-consuming, costly, and less accurate recycling process. Automated segregation in the proposed work uses Improved Deep Convolutional Neural Network (DCNN). The dataset of 2 class category with 25077 images is divided into 70% training and 30% testing images. The performance metrics used are classification Accuracy, Missed Detection Rate (MDR), and False Detection Rate (FDR). The results of Improved DCNN are compared with VGG16, VGG19, MobileNetV2, DenseNet121, and EfficientNetB0 after transfer learning. Experimental results show that the image classification accuracy of the proposed model reaches 93.28%.
Bhagwati Sharan and R Manjula
IEEE
Prolonged network operations are crucial for any Wireless Sensor Network (WSN) based applications such as health care, military, industrial, etc. The fixed energy and the restrained communication range of the sensor nodes (SNs) make it challenging to achieve a longer life of network operations. Generally, the data collection and direct message transmissions to the base station (BS) consume most of the SNs’ energy resulting in a shorter network lifetime. Reduction in the energy expended on these data transmissions leads to an improvement in the WSNs’ network lifetime. Clustering approaches are employed to mitigate this issue. The cluster head (CH) in each cluster collects the data from the SNs and forwards it to the BS. Proper CH node selection is vital in achieving enhanced network lifetime. Therefore, in this work, a new clustering technique is proposed that probabilistically selects the best CHs for each cluster by considering the residual energy of each SN. The suggested technique has been compared with the existing LEACH scheme and the experimental outcomes show the suggested technique has improved the network lifetime by an average of 8.32% and 14.23% in terms of the first and the last node dead respectively.
Tanushree Gupta, Gaurav Kumar Pandit, Ashutosh Kumar, Himanshu Mishra, and Bhagwati Sharan
IEEE
Women's safety is one of the major concerns nowadays. Women are not safe in societies, public spaces, and elsewhere, subjected to a lot of unethical physical harassment at bus stations, markets, bus stands, and other public places. In this research, data from different locations and states in India are collected to find out the highest crime rate zones. So that this research can further be helpful to make small wearable devices and also the applications using that data and find out the safest possible routes and eases the traveling for women coming late at night from offices. So, in this research, data collected for crime towards women and using that data to find out the red zone areas in different states according to that different sensors will be used for the devices that can help to track location, can call the top 5 emergency contacts and spark generator can also at the same time help the women. Taking into consideration device will be made multi-purpose.
Megha Chhabra, Bhagwati Sharan, and Manoj Kumar
IOS Press
The users of mobile phone are exponentially increasing. The applications are developed every day in a variety of domains to enhance the Quality of User Experience (QoUE) along with utility determinants. The design of the mobile application impacts the QoUE. QoUE in mobile applications is a measure that describes the appropriateness of the purpose of the application and the need for user retention. However, the challenge is to identify, understand, focus and interconnect the variety of determinants influencing the QoUE based on mobile application design. These determinants are based on the diversity of users and the related functional needs, user-specific needs, and background functioning of the application. The modelling and analysis help mobile application developers to improve, increase and retain user engagement on the app based on improved QoUE. To do so, a qualitative analytical method is employed in the following steps. The first ever Fuzzy Cognitive Map (FCM) is proposed to show the causal-effect links of the interdependent determinants in mobile applications based on QoUE. In our model, the existence of relationships between determinants relies on a thorough literature review. The weight of these links is estimated by users of different ages and lines of work. This is performed by an empirical study based on a questionnaire filled by experts. The questionnaire is based on the formal utility and perceived QoUE-based topics. Finally, scenario-based analysis on formed FCM based on these inputs is performed. We show that small changes in cases using different direct determinants can be used to enhance QoUE. These changes can be studied before launching an application for the user, thereby limiting the need to rework the improvements based on QoUE and providing useful guidance for the possible increase in user base and behaviour change.
Pulkit Dwivedi and Bhagwati Sharan
IEEE
The identification of facial key-points is a difficult challenge in computer vision. Each person’s face has incredibly distinctive facial characteristics. The centres and borders of the eyes, the brows, the nose, and the lips are some of the key elements of the face. Thus, finding the facial key features in a particular face is the major aim of the facial key point detection process. This is a very challenging task. In this work, the initial step of the proposed methodology for identifying facial key-points includes a range of training data augmentations to improve the number of training instances and the generalizability of the proposed model. In the second phase, an Inception-based deep neural network model is proposed that accomplishes this task quicker by reducing the training time. While testing, we combined the images from different data augmentation approaches to enable averaged predictions. We evaluate the model design with and without data augmentation techniques based on their Mean Squared Error (MSE) scores on the test set. The research demonstrates that the MSE of the proposed Inception-based model is very low as compared to the other state-of-the-art methods.
Bhagwati Sharan, Anil Kumar Sagar, and Megha Chhabra
IEEE
Vehicular Ad-hoc Networks (VANETs) enable vehicle-to-everything communication, where vehicles disseminate messages to a Road Side Unit (RSU) periodically. All RSUs send data to a cloud or a central server for detection and analysis of traffic congestion situations on the roadways globally. The existing cloud computing approach is inefficient for analyzing massive amounts of data in a short amount of time while still meeting the needs of consumers. Due to its limited scalability, flexibility, and connectivity, conventional vehicular networks have several issues in resource placement and administration, affecting the quality of service (QoS) and severely affecting VANET services and entire network effectiveness. To address these issues, a novel architecture is known as edge computing - which allows the decentralization of data preprocessing from the clouds to the edge of the network - had been positioned to solve the issues that have arisen while employing cloud computing method. Edge computing is defined by its ability to implement with VANETs to calculate, store, and deliver delay-sensitive communications to vehicles on deadline. Less latency, network off-loading, and context-awareness are just a few of the benefits that it might bring to the global vehicular network (location, environment factors, etc.). Mobile edge computing (MEC), fog computing (FC), and cloudlet are the primary methods to edge computing that have been developed. This paper presents a survey on cloud and edge computing, a detailed comparison of the existing research, characteristics, requirements for enabling edge computing, and challenges.
Kiran Kumar Ravulakollu, Megha Chhabra, Bhagwati Sharan, Ruchi Agarwal, Ritu Dewan, and Mayank Goyal
IEEE
Mobile Cell phones are essential for many young people; however, such devices can adversely affect their mental health and well-being. The rapid development of mobile technology offers a choice of high-end options and improved immobility, which will lead to an increase in the prevalence of mobile device use, especially among young people. They usually develop an attachment to mobile phones, seek closeness to mobile phones, and experience stress during separation. In the Asian country alone, about 600 million sensitive phone users will be active by early 2021. By the end of 2018, it is predicted that 530 million people in India will be smartphone users. Mobile addiction problems are growing at associate around the world at an alarming rate. Problems due to mobile phone addiction like sleep disturbance, anxiety, stress, and, to a lesser extent, and depression require immediate action. In this paper, several symptoms, effects, and causes of smartphone addiction are summarized.
Bhagwati Sharan, Megha Chhabra, and Anil Kumar Sagar
IEEE
Vehicular Ad-hoc Networks (VANETs) is a very fast emerging research area these days due to their contribution in designing Intelligent transportation systems (ITS). ITS is a well-organized group of wireless networks. It is a derived class of Mobile Ad-hoc Networks (MANETs). VANET is an instant-formed ad-hoc network, due to the mobility of vehicles on the road. The goal of using ITS is to enhance road safety, driving comfort, and traffic effectiveness by alerting the drivers at right time about upcoming dangerous situations, traffic jams, road diverted, weather conditions, real-time news, and entertainment. We can consider Vehicular communication as an enabler for future driverless cars. For these all above applications, it is necessary to make a threat-free environment to establish secure, fast, and efficient communication in VANETs. In this paper, we had discussed the overviews, characteristics, securities, applications, and various data dissemination techniques in VANET.