Authors - Ketaki Bhoyar, Suvarna Patil Abstract - Fashion is the canvas of our identity. Fashion can be so inclusive, expressive and sustainable. In the growing landscape of fashion, every individual face an overwhelming array of choices. It becomes difficult to discover a personalized wardrobe that reflects their preferences, taste and needs. Traditional Fashion Recommendation Systems (FRSs) limits their ability to scale and adapt the ever-growing styles as they heavily rely on manual design. Around the world, a large number of users buy cloths online through the e-commerce websites. These websites primarily use recommender systems. Appropriate recommendations given by FRS helps to enhance user satisfaction and makes it more enjoyable and accessible. Artificial Intelligence (AI) tools have revolutionized FRS enabling them to consume beyond conventional methods by taking in contextual data, user preferences and visual content for recommendations with a more individualized suggestion. Recently, Generative Adversarial Networks (GANs) have emerged as a potent technique to enhance these systems by generating diverse fashion designs with high fidelity. In this paper, a systematic review of parameters used to evaluate FRS using Generative Algorithms is discussed. Various parameters to evaluate system performance and the recommendation quality are analyzed. Detailed analysis of the input parameters, to be considered to design the efficient AI based FRS (AI-FRS) is also presented. Along with this, research gaps are explored by surveying numerous review papers. This review will help in deciding the evaluation parameters to develop and examine more efficient AI based FRS.
Authors - Shobha K, Rajashekhara S Abstract - The Service Selection Board (SSB) evaluates candidates for admission to military services like the Indian Army, Navy, and Air Force through a rigorous five- to six-day selection process. This process assesses a candidate’s psychological and physical fitness, communication skills, and leadership qualities. Despite its importance, the low selection rate highlights a lack of preparation platforms for aspirants. Many candidates cannot afford offline coaching, and no comprehensive online platforms exist to simulate SSB tests. The proposed solution is an interactive online platform offering real-time test simulations, feedback, and guidance, replicating the SSB interview experience to enhance aspirants’ chances of success.
Authors - Payal Khode, Shailesh Gahane, Arya Kapse, Pankajkumar Anawade, Deepak Sharma Abstract - The COVID-19 pandemic has led to a widespread trend toward remote work, drastically altering the nature of the traditional workplace. While there are many advantages to working remotely, such as flexibility and less time spent traveling, there are also major cybersecurity risks. The inherent vulnerabilities in technologies used for remote work pose a persistent threat to cybersecurity. But social distancing measures imposed by the pandemic have made workers work from home, which has increased internet usage. These widespread modifications have been used by malicious hackers to launch extensive phone scams, phishing attacks, and other computer-based exploits. Organization have quickly embraced remote work without fully understanding the impact on cybersecurity. Because remote work policies have been widely adopted without first consulting cybersecurity experts or implementing comprehensive security measures, there are now more vulnerabilities. This study focuses on people because they are the weakest link in cybersecurity. It highlights how important it is to protect business and personal information when working from a distance. The study looks at the cybersecurity risks associated with changing employee behaviors during the transition to remote work in light of the COVID-19 pandemic. The aim of this research is to investigate the cybersecurity risks and challenges that companies and organizations encounter when workers change their work habits to work remotely during the COVID-19 pandemic.
Authors - Ananya Solanki, Leander Braganza, Aarol D’Souza, Sana Shaikh Abstract - For plant lovers, there has always been a barrier to accessing a wider variety of plants. This restriction is due to the absence of a dedicated marketplace to buy and sell plants. This platform includes an interactive Augmented Reality (AR) feature that enables users to visualize the plants they select in their selected environment. Furthermore, it utilizes location-based Air Quality Index (AQI) and recommends plants according to the user’s location. This Platform will educate the users about the plants by providing plant care tips.
Authors - Prajkta Dandavate, Ameya Badge, Mohit Badgujar, Aditi Badkas, Rutuja Badgujar, Orison Bachute, Vedant Badve Abstract - This paper presents a cutting-edge combination system designed to integrate personalized music recommendations with real-time face-based emotion recognition by using adaptive emotion-driven user interaction. The approach demonstrates how, given a continuously streamed video coming from a PC camera, advantage is taken to analyze emotions as the CNN feeds in user-defined emotions in the emotion categorization task and indicates that such categories of emotions have been quite accurately identified or classified up to around 65% into defined categories, say for example sadness, happiness, and many more. It detects emotions within a room in real time while online building up a playlist of music. The system remains smooth and adaptive, constantly readjusting the emotional responsiveness of the interaction, supported by a multi-threaded architecture. In addition to entertainment, the paper explores other applications in home automation, healthcare, and mental health as well as opportunities for emotion-driven content and advertisements that match the real-time emotional states of users. It brings to the foreground the prospects of machine learning and the possibility of real-time processing in creating deeply personalized, emotionally driven user experiences across diverse settings.
Authors - Shailesh Gahane, Payal Khode, Arya Kapse, Deepak Sharma, Pankajkumar Anawade Abstract - In Mozambique, in recent years, the construction sector has seen a lot of loss of life caused by accidents at work, mainly due to the lack of control of international standards for OHS, the production process and employee orientation. Risk analysis and management advocates that risks can be characterized by being partially known, changing over time and being managed in the sense that human action can be applied to change their form and/or the magnitude of their effect. The field of artificial intelligence (AI) is experiencing rapid growth and is increasingly integrating into various sectors, including healthcare, industry, education, and the workplace. Its overall objective is to develop an environmental, health, and safety management system integrating artificial intelligence (AI) and blockchain to prevent accidents, facilitate decision-making, and comply with international construction regulations at sites in Maputo, Mozambique. To achieve this goal, the system will focus on administration and legal compliance, education and training, safety and emergency
Authors - Swapnil M Maladkar, Praveen M Dhulavvagol, S G Totad Abstract - Blockchain technology has emerged as a powerful tool for secure, decentralized data management across various industries, but it faces significant scalability challenges due to the limitations of existing sharding methods. Traditional static sharding approaches often result in inefficient resource allocation, while adaptive sharding techniques can lead to increased complexity and delayed adjustments, hampering overall system performance. This paper proposes an innovative blockchain network management approach by integrating Long Short-Term Memory (LSTM) models with dynamic sharding. This system leverages predictive analytics to optimize real-time sharding adjustments, significantly enhancing blockchain performance. By addressing the shortcomings of both static and adaptive sharding methods, the proposed approach avoids the extra infrastructure and delays associated with Layer 2 solutions. Future research will focus on advancing LSTM techniques, integrating them with other optimization strategies, and testing in real-world scenarios to further enhance scalability and efficiency. This LSTM-integrated dynamic sharding method represents a significant step forward in blockchain network optimization, offering a more efficient and adaptable solution for contemporary blockchain applications. Experimental results reveal a 22% increase in transaction throughput and a 25% reduction in latency compared to conventional static sharding.
Authors - Prema Sahane, Anand Dhadiwal, Devvrath Datkhile, Harshal Deore, Atharva Shinde, Amruta Hingmire Abstract - The paper provides information about different healthcare applications that are built to develop the healthcare sector digitally with the help of modern technologies. It describes the need for making the particular application with its advantages and disadvantages. Though there are many health record management systems existing for electronic health record management, the accuracy and efficiency are not up to the level that society need. People find a lot of time wastage in maintaining the records manually. Also, Patients find it difficult to track their previous records. So, our system “Medicard” is an application for interaction between doctors, patients, and pharmacists. It is a multi-tasking application for all healthcare tasks like Centralized Storage of patient health records, Drug Analysis, Allergy Analysis, Online receipt generation, Community creation, Booking Doctor’s Appointments and Online Payment. It has three different interfaces for doctors, patients, and pharmacists.
Authors - Anant Chovatiya, Priyanka Patel Abstract - Attendance management holds significant importance for all organizations, serving as a determining factor in their success, whether they operate in educational institutions or the public and private sectors. Efficiently tracking individuals within the organization, including employees and students, is crucial for optimizing their performance. Managing employee attendance during lecture periods has become a challenging endeavor. The task of computing attendance percentages poses a significant challenge as manual calculations often result in errors and consume excessive time, leading to inefficiencies and time wastage. In response to the challenges posed by traditional paper-based practices in educational institutions, this paper introduces a digital solution for managing university lecture slots and attendance. The proposed system, named the "Speed Check system," aims to streamline faculty and student attendance processes through a mobile application, eliminating the need for manual recording and reducing paper consumption. Leveraging a cloud-based NoSQL database, real-time data synchronization ensures seam-less communication across users. The system offers distinct functionalities for Time Table Coordinators and Attendance Coordinators, facilitating efficient slot scheduling, modification, and attendance marking. Utilizing Flutter SDK and Firebase technology, the application provides a user-friendly inter-face and robust data protection. Future enhancements include role-based access control and advanced analytics for informed decision-making. Overall, this digital solution presents a significant stride towards optimizing academic administration and enhancing the effectiveness of attendance management in educational institutions.
Authors - Nirali Arora, Harsh Mathur, Vishal Ratansing patil Abstract - Achieving relevance in search results is difficult in today's complex information environment, particularly when single-algorithm ranking models find it difficult to account for a variety of user circumstances. In order to improve search relevancy in a variety of circumstances, this study presents a unified ranking strategy that integrates many algorithms. Hybrid system adapts dynamically to user intent and situational details by combining conventional models like BM25 and PageRank with cutting-edge neural techniques like BERT-based transformers and learning-to-rank algorithms. A key component of this strategy is a context recognition mechanism that continuously evaluates user history, query type, and behavioural patterns to fine-tune relevance score according to the particular requirements of every search context. This method, called Contextual Rank, combines algorithmic scores to prioritize relevance, enabling more flexibility and response to user demands. Here presented about the theoretical ramifications, covering problems like scalability and processing needs as well as gains in relevance. The benefits of unified ranking models are highlighted in this paper, opening up new avenues for contextual optimization in recommendation systems and search engines and paving the way for improved user experiences across a range of search settings.
