Vivian Lobo
@sjcem.edu.in
Assistant Professor, Department of Artificial Intelligence and Machine Learning
St. John College of Engineering and Management, Palghar
RESEARCH, TEACHING, or OTHER INTERESTS
Computer Engineering, Computer Science, Computer Networks and Communications, Artificial Intelligence
Scopus Publications
Scholar Citations
Scholar h-index
Scholar i10-index
Scopus Publications
Amit Pal, Ashishkumar Tiwari, Amol Singh, Manishsingh Tanganiya, Rashmi Bhat, and Vivian Brian Lobo
IEEE
Today, in this technical world, where all subjects are changing from traditional approach to digital approach, extended reality (ER) has played an important role in supporting this change. One of the efficient parts of ER is virtual reality (VR). VR can be defined as bringing a virtual world or fictitious world into real life where a user can explore the world in his/her own home space without any interference. VR can be well applied in various fields such as education like viewing solar space in school using VRHeadset in immersive way so that students can understand the concept clearly, construction and industrial regions like designing the building in virtual space before starting the actual development process so that the engineer have proper idea for development without much loss, simulations including training regions, gaming industries for entertainment, tourism, virtual tour, etc. Considering virtual tour as one of the test cases of VR, creating a virtual tour of St. John College of Engineering and Management (SJCEM) campus would be much preferable for visitors so that students can not only get an immersive experience of the campus but also can explore different sections of the campus without any interruptions using VR at their own home space. This study aims to provide a virtual tour of SJCEM campus to a user whose can not only explore the campus but also take an in and out immersive experience.
Omkar Nikam, Shreyash Singh, Swaraj Patil, Suryakrishnan Nair, Aniket Raut, Vivian Brian Lobo, Shraddha More, and Ronald Melwin Laban
IEEE
Suicides are a serious matter in the modern society. In India, suicide ideation/attempt is noticed to be one of the most common trends amongst youth. With each succeeding year, suicide rates are increasing in India. The southern and eastern states of India exhibit trends of high suicide rates. Current methods of suicide ideation/attempt detection include clinical visits, online counseling, etc. This study aims to propose an automated system that focuses on envisaging suicide rates of vivid states of India. Moreover, the proposed system forecasts an individual’s risk of suicide with the help of a questionnaire filled and previously observed parameters. Machine learning techniques such as linear regression and decision-trees are used, which achieved 85.98% and 90.476% accuracy, respectively.
Swaraj Patil, Omkar Nikam, Suryakrishnan Nair, Aniket Raut, and Vivian Brian Lobo
IEEE
Efficient supply chain management is a laborious task in any company, but in food industry, there is extra complexity and risk of compromising supply chain in food waste management, which may directly affect consumer safety and the shelf life of goods. One possible solution is to use blockchain technology to improve the integrity, security, and transparency of food supply chain. This study intends to present a blockchain-based transparent traceability food supply system that would follow the passage of boxed leftovers from restaurants through nongovernmental organizations and lastly to the underprivileged. Furthermore, the system employs a QR code and an NFC tag to acquire access to secure data pertaining to food packaging. This can help to reduce fraud, theft, and counterfeiting, as well as improve the efficiency of leftover food supply chain.
Vivian Brian Lobo and Madhuri Rao
IEEE
The significance of blockchain technology and its applications in commodity markets have been gaining not only recognition but also popularity over the past few years. One of the main reasons behind the acceptance of blockchain technology in commodity markets is efficiency. The concept of centralization has been radically transformed because of blockchain technology. Decentralization, tamperproof nature, transparency, peer-to-peer (p2p) exchange, amongst others are some of the characteristics of blockchain technology that help in tackling issues faced by commodity markets. This study aims to present an empirical study of commodity market-based systems that enable the use of blockchain technology in various business processes.
Nidhi Agrawal, Himanshu Kushwaha, Saujanya Shetty, and Vivian Brian Lobo
IEEE
There are a large number of fake products in the market being manufactured, distributed, sold, and purchased. Here, the meaning of fake products is products that are not from registered or branded sellers and manufacturers. Fake products are actually gaining more importance in the market because of low cost and similarity of them with original products. In this study, genuinity of fake products is determined to help users so that they can distinguish between fake and original products with the help of blockchain technology and cryptocurrency. QR codes will help users to scan and acquire overall information of a product such as basic details, genuinity, ownership, and seller information.
Neel Bhatt, Pratiksha Laldas, and Vivian Brian Lobo
IEEE
We simplify almost everything in our lives today by automating tasks. When driving, we often miss signs on the side of a road because our attention is focused on the road. This is hazardous to us and to people around us. If a driver is alerted without having to shift his/her attention, then this problem can be avoided. Traffic sign detection and recognition (TSDR) plays a major role in this as it detects and recognizes traffic signs, thereby alerting a driver if any traffic signs are approaching. By doing so, besides the safety on road being ensured, the driver will feel more relaxed when navigating tricky or unfamiliar roads. Oftentimes, it is difficult to comprehend signs and other warnings. Using this approach, drivers will not have to struggle with translating signs. This study aims to propose a model for TSDR using deep learning wherein convolutional neural networks are used for recognizing traffic signs using a hybrid dataset that comprises German traffic sign recognition benchmark dataset from Kaggle and a self-created Indian traffic sign dataset. In addition, the proposed model is trained on both datasets individually so that the output from the proposed model can be compared with the output of existing models. Experimental results revealed that the proposed model attained an accuracy of 95.45% for hybrid datasets and for the Indian dataset alone 91.08%, whereas for the German dataset alone, accuracy yielded was 99.85%.
Aditya Vijaykumar Singh, Ashwin Omprakash Tiwari, Shreyash Sanjay Singh, and Vivian Brian Lobo
IEEE
In India, as per records, it has been observed that with each succeeding year, criminal activities are surging. Crime is an act executed against the law of constitution and is a menace to our society. Thus, crime and criminals need to be monitored conveniently. Unlike USA or other countries, in India, criminal records are not public, which means neither private organizations nor common people can access criminal records for employee verification. A centralized system, which gives access to criminal data, would be a feasible solution to this problem. This study aims to develop a Flutter-based blockchain-enabled permissionless system that would help in the remote access of criminal records. The system would be developed using Ethereum, which offers vivid features such as truffle for enabling Ethereum virtual environment, Ganache for personal testing of smart contracts, InterPlanetary file system for encrypted storage of data and easy tracking of files, and MetaMask, which acts as a digital wallet.
Ritik Dhedia, Nixon Paliakkara, Vivian Brian Lobo, Deepak Gupta, and Vaibhav Sharma
IEEE
This study aims to help farmers by using open source software that employs machine learning and hyperspectral images to analyze farm characteristics, which include crops, soil, and climate. This study makes use of two datasets, i.e., 270100 images from LANDSAT 8 and classified images from MODIS dataset provided by Google Earth Engine to classify land type, which helps in detecting farms in the future. Random forest algorithm was used as a classifier for multiclass hyperspectral data. Training the model acquired an overall accuracy of 0.997 that helped to determine the type of land in a geographical area. This paper conveys the first model built by us from various other models that are planned to develop. The data from our research work is conveyed to a farmer by means of a web application, which is built using a Spring framework, Grafana, JvaScript, and several other web technologies.
Devika Babrekar, Darsh Patel, Sachin Patkar, and Vivian Brian Lobo
IEEE
Our identity as a human being is determined by the documents, not by appearance or physicality. The most important thing to prove the identity of humans is to show a government-issued document. Generally, from birth to death humans are recognized by documents because they are born with a birth certificate and they die with a death certificate. The main problem with these documents is that, they can be falsified or manipulated by others. Moreover in this digital era, they are stored in a centralized manner, which is prone to a cyber threat. This study aims to develop a blockchain environment to create, verify, and securely share documents in a decentralized manner. With the help of bigchainDB, interplanetary file system (IPFS), and asymmetric encryption, this research work will prototype the proposed solution called blockchain-based digital locker, which is similar to the DigiLocker released by the Department of Electronics and Information Technology (DeitY), Govt. of India. BigchainDB will help in treating each document as an asset by making it immutable with the help of IPFS and asymmetric encryption, where documents can not only be shared but also verified.
Jayesh Patil, Vaibhav Patil, Dhaval Walavalkar, and Vivian Brian Lobo
IEEE
Road accidents are a major cause of fatalities in India and other nations too. Fatality rate in developing nations is very high due to various aspects. In the past, it was assumed that road accidents and fatalities cannot be avoided, but now with this tech era, everything is almost becoming possible. Machine learning (ML) is used to analyze various algorithms through experience and improve results. It includes three major types of learning techniques, namely supervised, unsupervised, and reinforcement learning. Our study focuses on reducing mortality rate by setting up a prediction model by means of an unsupervised learning technique, i.e., k-means clustering, which analyzes road accidents by taking into consideration various aspects like potholes on roads, sharp turns, and weather conditions and then provides suitable and precautionary measures to avoid mishaps by representing it on map and creating an intelligible model for everyone. The predicted model achieved an accuracy of 81%.
Rishabh Bhatangar and Vivian Brian Lobo
IEEE
This study aims to provide a novel concept of multihop concurrent communication between blockchain network peers to facilitate big data sharing via threads in a multithreaded environment to reduce effective transmission time required for sharing a large file. In a general client–server model, the entire load of sharing and synchronization of files is split among a service provider and a client. A three-tier model too provides the same rate of speed irrespective of the number of nodes a current node is connected with, provided there is no network congestion. This paper focuses on how decentralization might help to effectively overcome such problems. Decentralization would increase the number of workers that will redirect data from a server node to a client node and blockchain would maintain security when data is being shared. The proposed model can be used to transfer big files in a relatively less span of time. Moreover, the study elaborates on design implications and considerations a user must understand to effectively use this model for achieving the holy grail to acquire better results when implemented on his/her systems.
Jayesh Patil, Mandar Prabhu, Dhaval Walavalkar, and Vivian Brian Lobo
IEEE
Accidents through roadways have been a great threat to developed as well as underdeveloped countries. Road accidents and its safety have been a major concern for the world, and everyone is trying to handle this since years. Road traffic and reckless driving occur in every part of the world. Because of this, many pedestrians are affected too. With no fault, they become victims. Many road accidents occur because of numerous factors like atmospheric changes, sharp curves, and human faults. Injuries caused by road accidents are major but sometimes imperceptible, which later on affect health too. This study aims to analyze road accidents in one of the popular metropolitan cities, i.e., Bengaluru, through k-means algorithm and machine learning by scrutinizing accident-prone or hotspot areas and their root causes.
Arbaz Khan, Rishabh Bhatnagar, Vinit Masrani, and Vivian Brian Lobo
IEEE
Intensifying requests for energy is preceding towards renewable solar energy integration with nonrenewable energy resources. Unlike other nonrenewable energy resources, solar energy is recurrent. Effectual utilization of spontaneously available energy accurate solar power forecasting is essential. This study aims to predict solar power through deep neural networks (DNNs) and various machine learning (ML) techniques on a solar dataset, namely linear regression, support vector regression, random forest, etc. The dataset that is used contains solar power energy extracted every five minutes. Moreover, a comparative study is carried out between DNNs and ML techniques, which helps in crafting suitable decisions to select appropriate forecasting and prediction techniques.
Vivian Brian Lobo, Jetso Analin, Ronald Melwin Laban, and Shraddha S. More
IEEE
Owing to enlarged digital data obtainability and artificial intelligence (AI) progressions, there are quite a few occasions that can be reconnoitered in healthcare. Deep learning (DL) and inductive transfer practices are turning healthcare data— such as phantasmagorias and videotapes—into powerful data sources for predictive analytics. At the present time, patients fail to have entree to his/her individual medicinal records and hang around ignorant of data importance or prominence. This paper directs to offer a gestalt of AI and blockchain and exhibit a roadmap for a blockchain-assisted decentralized bionetwork of private healthcare data to expediate new methodologies to drug discovery and precautionary healthcare. A protected and crystal-clear disseminated marketplace of personal data by means of blockchain and DL technology will circumvent challenges faced by authorities of a given healthcare system and restore custody all across private records that includes medicinal documents back to humans. It also proposes a novel type of utility cryptotoken named LifeCoin, which can be produced through the stationing of data on the blockchain-assisted open market to streamline transactions and expedite inventive reward schemes.
Umang Mavani, Vivian Brian Lobo, Aditi Pednekar, Naomi Christianne Pereira, Rupesh Mishra, and Nazneen Ansari
Springer Singapore
Data mining (DM) is used to analyze and classify data and identify hidden patterns stored in a data warehouse in an attempt to predict future trends, which are quintessential to knowledge discovery and provide tremendous support not only to the world of business but also to that of academia. There are various open-source and freely available software tools such as Weka, R, and Orange as well as programming languages like Python used for DM. This study focuses on comparing the performance of these tools by performing Naive Bayes classification on student placement data. Percentage of marks scored by students in S.S.C. and H.S.C. examinations and their engineering aggregate were inputs to the tools. Moreover, the tools were trained and tested to decide whether a student would be placed or not. Comparative analyses of the tools were done to determine which tool was able to provide the highest prediction accuracy on student placement data.
Naomi Christianne Pereira, Umang Mavani, Aditi Pednekar, and Vivian Brian Lobo
IEEE
Students attaining higher schooling in specialized courses are often unaware of grading systems that are followed because of which they find it challenging to cope up with their ongoing course. They are unable to extemporize on their study plan and fail to accomplish better scores. There could be innumerable reasons (both personal and professional) like health concerns; family occasions; involvement in extracurricular activities such as technical, cultural, sports, among others-which ultimately leads to inadequate scores at the end of a semester of an academic year. There felt a need to develop a system that can estimate a range of scores for internal assessment tests (IATs) and end semester examinations (ESEs) for students so that they can prepare accordingly for IATs and ESEs and attain their desired pointer in a semester. There are several educational systems trying to predict students' grades or scores, but they face certain limitations like accuracy, response time, and error handling. This work aims to overcome the abovementioned limitations by developing a state-of-the-art estimation system that calculates semester grade point average by considering students' IAT marks as input and forecasts minimum and maximum marks that need to be scored in ESEs based on a selected pointer. Moreover, the system is capable of specifying ways to achieve the desired pointer without the inclusion of IAT marks. The developed system will undeniably be beneficial for both students and educational institutions wherein a student can beforehand calculate his/her academic performance and institutions can timely and precisely monitor students' growth leading to future betterment.
Mrunmayi Patil, Vivian Brian Lobo, Pranav Puranik, Aditi Pawaskar, Adarsh Pai, and Rupesh Mishra
IEEE
Diseases that are associated with the way a person or group of people live are known as lifestyle diseases. Healthcare industry collects enormous disease-related data that is unfortunately not mined to discover hidden information that could be used for effective decision making. This study aims to understand support vector machine and use it to predict lifestyle diseases that an individual might be susceptible to. Moreover, we propose and simulate an economic machine learning model as an alternative to deoxyribonucleic acid testing that analyzes an individual's lifestyle to identify possible threats that form the foundation of diagnostic tests and disease prevention, which may arise due to unhealthy diets and excessive energy intake, physical dormancy, etc. The simulated model will prove to be an intelligent low-cost alternative to detect possible genetic disorders caused by unhealthy lifestyles.
Vivian Brian Lobo and Shamsuddin S. Khan
IEEE
A sensor is a device that senses, perceives, or distinguishes and responds to an input from a physical location and provides an output that is human decipherable. As obesity levels all over the world are reaching the sky, especially in the U.S., China, and India, the use of sensors can help in managing rising obesity levels. There are many traditional dietary assessment methods such as doubly labeled water, 24-hour recall, paper-based food diaries, and food frequency questionnaires used for managing and reducing obesity, but they have many limitations. Therefore, modern dietary assessment methods such as acoustic-, image-, motion-based, unobtrusive, and multimodal methods have been developed. Devices that use these modern dietary assessment methods are automated wrist motion tracking system, WearSens necklace, and automatic ingestion monitoring system. This study provides a review of such devices that help in reducing obesity for better living in humans.
Vivian Brian Lobo and Nazneen Ansari
IEEE
Twitter, a well-liked online social networking site, facilitates millions of users on a daily basis to dispatch and orate quick 140-character notes named tweets. Nowadays, Twitter is cogitated as the fastest and popular intermediate of communication and is used to follow latest events. Tweets pertaining to a specific event can be effortlessly found using keyword matching, but there are numerous tweets that are likely to contain information that is semantically identical. Moreover, there exist many systems for recapitulating tweets related to a particular event, but they have numerous limitations and are unable to provide accurate results. This work aims to overcome the limitations of existing systems by developing a system for recapitulating tweets using graph-based clustering.
Vivian Brian Lobo, Nazneen Ansari, and Rajkumar K. Shende
IEEE
Twitter, a well-liked online social networking site, facilitates millions of users on a daily basis to dispatch and orate quick 140-character notes named tweets. Nowadays, twitter is cogitated as the fastest and popular intermediate of communication and is used to follow latest events. Tweets pertaining to a specific event can be effortlessly found using keyword matching, but there are numerous tweets that are likely to contain information that is semantically identical. Moreover, there exist many systems for recapitulating tweets related to a particular event, but they have numerous limitations and are unable to provide accurate results. This study aims to overcome the limitations of existing systems by proposing a system for recapitulating tweets using graph-based clustering.
Vivian Brian Lobo, Blety Babu Alengadan, Sehba Siddiqui, Annies Minu, and Nazneen Ansari
IEEE
Traveling salesman problem (TSP) is studied as a combinatorial optimization problem—a problem that attempts to determine an optimal object from a finite set of objects—which is simple to state but difficult to solve. It is a nondeterministic polynomial-time hard problem, hence, exploration on developing algorithms for the TSP has focused on approximate methods above and beyond exact methods. The mission in the TSP is to determine the shortest (optimal) tour when a salesman travels across many cites. A major challenge is that the salesman must be able to minimize entire tour length. The solution to the TSP experiences eclectic applicability in various fields and thus advances the need for an effectual solution. There have been exertions heretofore to provide time efficient solutions (i.e., exact as well as approximate) for the TSP. Dynamic programming is an effective and powerful method that could be used to solve the TSP. Generally, for solving the TSP, a unidirectional path is provided (i.e., whether the salesman travels from city A to B or city B to A) in any input graph, and so, it becomes easier in determining the shortest tour. However, in our study, we have considered a situation where no directions are specified (i.e., the salesman can travel both from city A to B and from city B to A) in an input graph, and for such a graph (i.e., a bidirectional graph), we will determine the shortest tour using dynamic programming.
Vivian Brian Lobo and Nazneen Ansari
IEEE
The complex construction of online educational systems lies within three key activities, i.e., design, execution, and appropriate post-implementation assessment. However, there is inadequate knowledge with regard to these activities. Effective execution of these three activities demands the use of design and educational models to obtain time proficiency, cost, and high educational quality. The use of online educational systems would benefit from an organized approach to design, execution, and evaluation of students. Therefore, this study proposes a general design of both a model and framework for improving online educational systems for teachers as well as students by taking into consideration accurate assessment and effective evaluation of the learning process. In this study, we use a local area network-based connection for creating a virtual classroom that comprises both audio and video conferencing, which can be used by students as well as teachers. Moreover, we include other features such as moderated online chat between students and a teacher, resource sharing, questioning, survey, feedback, and query posting by students during the unavailability of a teacher. Moreover, we implement audio and video conferencing-based e-learning via real-time transport protocol.