@nirmuni.ac.in
Assistant Professor Computer Science and Engineering
Institute of Technology, Nirma University, Ahmedabad
Scopus Publications
Scholar Citations
Scholar h-index
Scholar i10-index
Pooja Chaturvedi, Kruti Lavingia, and Gaurang Raval
Springer Science and Business Media LLC
Vipul Narayan, A. K. Daniel, and Pooja Chaturvedi
Springer Science and Business Media LLC
Pooja Chaturvedi and A. K. Daniel
Springer Science and Business Media LLC
Henil Shalin Panchal, Nilesh Kumar Jadav, Pooja Chaturvedi, Rajesh Gupta, and Sudeep Tanwar
IEEE
The rapid advancement of digital technologies allowed people to interact with different social media platforms for communicating and sharing their feeds with their friends and communities. However, sharing and posting on social media are imperiled by various privacy-related issues, such as tampering with the uploaded images. Malicious individuals use manipulated images to exploit users and spread false information. Motivated by the aforesaid issue in this paper, we proposed a framework that can effectively detect manipulated images on social media platforms, specifically tackling image-based ransomware attacks. The proposed framework combines several image processing algorithms, including the Structural Similarity Index, Correlation Coefficient, Image Histogram Comparison, and Image Feature Comparison algorithms, to detect the manipulated images. Through experimentation and performance analysis, it is determined that the proposed framework has varying sensitivity levels to induced noise and manipulation. The results demonstrated the effectiveness of the proposed framework in detecting manipulated images, with the SSIM achieving the highest accuracy and the histogram comparison algorithm being the fastest.
Meet Mavani, Pooja Chaturvedi, and Swati Manekar
IEEE
One of the important purposes for machine learning is to analyze the data and use that data in order to forecast data and many times especially in time series data, forecasting becomes very hard. The need of analyzing and forecasting traffic flow has always been emphasized. Government can collect data and use the data for better traffic management and avoid many traffic congestions. The data collected can then be analyzed and used for traffic volume forecasting which can save a huge amount of time lost in commute. It may also save lives by allowing emergency vehicles, such as ambulances, to get to hospitals more quickly. The problems stated above need a solution where it becomes at most necessary to forecast the traffic volume in order to prepare for the same. In this comparative analysis four different models are used for predicting traffic flow. The models include ARIMA, FbProphet, XgBoost and LSTM. These selected models have been used in order to forecast the hourly traffic volume. The dataset for traffic flow analysis is taken from UCI machine learning repository. The performance is then evaluated on the basis of several parameters such as Root Mean Square Error(RMSE), and Mean Absolute Percentage Error(MAPE). The XgBoost model performs the best as it achieves the lowest MAPE among all the considered models. The LSTM model performs better than the other models considered for the study.
Faizal Kureshi, Dhaval Makwana, Umesh Bodkhe, Sudeep Tanwar, and Pooja Chaturvedi
Wiley
Pooja Chaturvedi, A.K. Daniel, and Vipul Narayan
Chapman and Hall/CRC
Pooja Chaturvedi, A. K. Daniel, and Umesh Bodkhe
Springer International Publishing
Pooja Chaturvedi and Ajai Kumar Daniel
IGI Global
Wireless sensor networks (WSNs) have attracted great attention because of their applicability in a variety of applications in day-to-day life such as structural monitoring, healthcare, surveillance, etc. Energy conservation is a challenging issue in the context of WSN as these networks are usually deployed in hazardous and remote applications where human intervention is not possible; hence, recharging or replacing the battery of sensor nodes is not feasible often. Apart from energy conservation, target coverage is also a major challenge. Scheduling the nodes to exist in active and sleep modes is an efficient mechanism to address the energy efficiency and coverage problem. The chapter proposes an ARIMA model-based energy consumption prediction approach such that the set cover scheduling may be optimized. The chapter compares the efficiency of several ARIMA-based models, and the results show that the ARIMA (0,1,2) model provides best results for the considered scenario in terms of energy consumption.
Pooja Chaturvedi and A. K. Daniel
Springer Science and Business Media LLC
Kruti Lavingia, Mihirsinh Vaja, Pooja Chaturvedi, and Ami Lavingia
IEEE
The goal of this paper is to design an automated system model to monitor the violation of traffic rules, specifically the number of people sitting on a two-wheeler. Typically, in areas near the security offices, people tend to follow the rules, but in areas where no one is watching, people violate the rules. In our case of an organizational campus, if there are three people traveling on a two-wheeler but when they encounter a security guard, one of the persons gets down and walks ahead of the guarded area and then again sits back on the vehicle. In such cases, efficient methods are required to monitor the violation of specified traffic rules without human intervention. For the above-mentioned challenge, a deep learning-based solution is provided where the process starts with object recognition using YOLOv3 (You Only Look Once) model, using which a person sitting on any particular vehicle is identified based on a minimum threshold distance. Also, for the distance calculation, a depth estimation algorithm which helps us in finding the 3-D distance between objects from a 2-D image is implemented. Moreover, the number plate of the vehicle violating the above-mentioned rule is identified for easy identification of the person violating the rule. The proposed approach is implemented on a real time video streaming dataset. The simulation results show the efficiency of the proposed approach in terms of accuracy, precision and recall as 91%, 86% and 94% respectively.
Pooja Chaturvedi and Ajai K. Daniel
Bentham Science Publishers Ltd.
Background: Target coverage is considered a significant problem in the area of wireless sensor networks, which usually aims at monitoring a given set of targets with the required confidence level so that network lifetime can be enhanced while considering the constraints of the resources. Objective: To maximize the lifetime of the sensor network and minimize the overhead involved in the scheduling approach, such that the pre specified set of targets is monitored for longer duration with the required confidence level. Methods: The paper uses a fuzzy logic system based on Mamdani inference in which the node status to remain in the active state is determined on the basis of coverage probability, trust values and node contribution. The rule set for determining the set cover is optimized by using the rough set theory, which aims to determine the node validity for the trust calculation. Results: The results show that the proposed approach improved the network performance in terms of processing time, throughput and energy conservation by a factor of 50%, 74% and 74%, respectively, as compared to the existing approaches. Conclusion: The paper proposes a scheduling strategy of the nodes for target coverage as an enhancement to the Energy Efficient Coverage Protocol (EECP) on the basis of rough set theory. The rule set for determining the set cover is optimized by using the rough set theory so that the network performance is improved in terms of the processing time, throughput and energy consumption.
Pooja Chaturvedi and Ajai Kumar Daniel
IGI Global
Wireless sensor networks have gotten significant attention in recent times due to their applicability in diverse fields. Energy conservation is a major challenge in wireless sensor networks. Apart from energy conservation, monitoring quality of the environmental phenomenon is also considered a major issue. The approaches that addressed both these problems are of great significance. One such approach is node scheduling, which aims to divide the node set into a number of subsets such that each subset can monitor a given set of points known as targets. The chapter proposes a priority coding-based cluster head selection approach as an extension of the energy efficient coverage protocol (EECP). The priority of the nodes is determined on the basis of residual energy (RE), distance (D), noise factor (N), node degree (Nd), and link quality (LQ). The analytical results show that the proposed protocol improves the network performance by reducing the overhead by a factor of 70% and hence reduces the energy consumption by a factor of 70%.
Pooja Chaturvedi and A. K. Daniel
Springer Singapore
Namrata Singh, A. K. Daniel, and Pooja Chaturvedi
IEEE
Human face is an important object in an image database due to its unique features (eyes, mouth, nose etc.) in every human being. The detection & recognition of a face in an image using template matching is one of the profound research interest in the field of image processing. Various approaches have been proposed in the literature to extract the visual facial features based on texture, color, shape, sketch & pose variance etc. for face detection in color images. This paper describes an approach of face detection technique that includes major characteristics such as lightening compensation based on luminance (Y) & chrominance (Cr), Color segmentation, skin-tone statistics & eye-mouth region computation. A template matching algorithm using cross correlation method for locating & recognizing a face has been applied on various face candidates to match the template with right face candidate. Thus, the presented work is divided into three steps: Face detection, Computation of template matching & Face recognition. The performance of given approach has been evaluated on the basis of run time & accuracy. The simulation result shows that the defined model is efficient in terms of accuracy which is 81% and the false alarms are reduced.
Pooja Chaturvedi and A. K. Daniel
IEEE
Wireless sensor networks have achieved the substantial research interest in the present time because of their unique features such as fault tolerance, autonomous operation etc. The coverage maximization while considering the resource scarcity is a crucial problem in the wireless sensor networks. The approaches which address these problems and maximize the network lifetime are considered prominent. The node scheduling is such mechanism to address this issue. The scheduling strategy which addresses the target coverage problem based on coverage probability and trust values is proposed in Energy Efficient Coverage Protocol (EECP). In this paper the optimized decision rules is obtained by using the rough set theory to determine the number of active nodes. The results show that the proposed extension results in the lesser number of decision rules to consider in determination of node states in the network, hence it improves the network efficiency by reducing the number of packets transmitted and reducing the overhead.
Pooja Chaturvedi and A. K. Daniel
IEEE
Wireless sensor network is a crucial research topic among the researchers because of the significant advances in the wireless transmission technology. Wireless sensor network is an infrastructure less network consisting of several small size and low cost devices possessing the sensing, computational and wireless communication abilities. There are several energy management, routing and data dissemination approaches proposed in the literature. The major emphasis however has been given to the routing protocols which differ on the basis of network organization and application requirements. Coverage is emphasized as major concern in the area of sensor networks which determines the degree and duration of sensing quality. The paper proposes a hidden markov model based node status prediction approach which determines the optimal sequence of states of the nodes such that all the targets are monitored with the desired coverage level. The trust concept is also incorporated to overcome the dynamism of the network conditions. The results obtained by using the dynamic programming mechanism and hidden markov model are compared.
Pooja Chaturvedi and A.K. Daniel
Inderscience Publishers
Pooja Chaturvedi and A. K. Daniel
IEEE
Wireless communication technology and the electronic systems advancements have laid the foundation of the wireless sensor network which achieves the growth of low-cost, low power and multi-functional sensor nodes which can be deployed on a single chip to communicate over a short range. The sensor node consists of sensing elements, actuator and a limited source of energy as battery. The major problem in wireless sensor networks is the efficient energy utilization and to keep the network in a good monitoring quality as long as possible. Coverage is a challenging problem in context of the sensor networks and is considered as a measure of QoS. The coverage problem is basically concerned about how well and how long the sensor nodes monitor the given target region. The most efficient way to ensure coverage, while preserving the energy resources is to schedule the nodes between active and sleep modes. Trust mechanism are incorporated for the uncertainties introduced by the environment in which they are deployed. The proposed protocol is an energy efficient coverage preserving protocol, which determines a set of nodes to keep in active state such that every target region is monitored by at least a sensor node. The confidence level of the node determines the level of data sensing and transmission to the base station.
Pooja Chaturvedi and A. K. Daniel
Springer International Publishing
Pooja Chaturvedi and A. K. Daniel
ACM Press
Maintaining the desired coverage level and reliability is a major challenge in the field of wireless sensor networks (WSNs). WSN consists of number of resource constrained and unreliable sensors. In this paper we have proposed a node scheduling protocol based on trust model, which reduces the energy consumption by putting the sensor nodes in the sleep mode. To improve the lifetime we have compared the performance of proposed protocol with Trust Based Probabilistic Coverage Algorithm (TBCA). The proposed Energy Efficient Coverage Protocol (EECP) provides the performance improvement in terms of network lifetime and reliability of the data transmitted to the base station.
Pooja Chaturvedi and A.K. Daniel
IEEE
The technological advancement in the field of wireless communication and electronic systems have provided the opportunity for the wireless sensor network to achieve the development of low-cost, low power and multi-functional sensor nodes which can be placed on a single chip and can communicate over a short distance. The critical issue faced by wireless sensor networks is the efficient energy utilization and is to keep the network alive as long as possible. The coverage problem is basically defined as how well and for how long the sensor nodes are able to monitor the given targets. The most efficient and effective way to ensure coverage, while preserving the energy is scheduling the nodes to remain in active or sleep modes. Trust metrics are incorporated in the networks to tackle the uncertainties present in the environment. In this paper, we propose an energy efficient target coverage preserving protocol, which determine a set of nodes to keep in active state such that every target region is being monitored by at least one sensor node and the data sensed by these nodes satisfies the predefined confidence level. In the proposed protocol, observer nodes monitor the neighborhood nodes to calculate the trust level. The nodes having the lower coverage probability are termed as observer nodes and are used to determine the trust value of the nodes, which is a convenient option as it reduces the communication overhead. The performance of the network is improved by incorporating the aggregation mechanism based on the link stability/availability and the residual energy of the nodes.
Pooja Chaturvedi and A. K. Daniel
IEEE
The existence of Internet, communications and information technologies, combined with recent Micro Electro Mechanical System advances, is leading the way for a new generation of low-cost sensors and actuators, which aim to achieve a high degree of spatial and temporal resolution and accuracy. The autonomous nature of sensor networks makes it of significant importance in various hazardous application such as battlefield surveillance, disaster management, structural health monitoring etc. Besides the numerous capabilities there are several issue which can significantly degrade the network functionality such as energy and resource constraint, holes problem etc. The problem of network holes is an important issue to consider. In this survey paper, we aim to provide the detailed overview of sensor holes, its causes, impact on the network performance and challenges related to this problem. We would also discuss some approaches which have been provided to overcome this problem. In this paper we aim to propose an approach which can detect and recover the coverage hole, which can result in the disconnection of the sensor networks.