Manuscripts

Recent Papers

Research Paper

IoT-Enabled Smart Home Gardening System: An Innovative Approach to Water Conservation and Plant Care

This paper introduces an IoT-based Smart Home Gardening System aimed at efficient water management and optimised plant care. The system leverages advanced soil moisture, temperature, and environmental sensors to monitor conditions and dynamically adjust water distribution using a mobile application. This integration of IoT technology supports water conservation, healthier plant growth, and sustainable gardening practices. The study highlights its potential applications for indoor, outdoor, and urban gardening spaces, focusing on scalability, affordability, and user-friendly features.

Published by: Debjyoti Mukhopadhyay

Author: Debjyoti Mukhopadhyay

Paper ID: V11I5-1149

Paper Status: published

Published: September 16, 2025

Full Details
Research Paper

Real-Time Football Player and Ball Detection System Using YOLO Architecture for Automated Sports Analytics

This paper presents a comprehensive AI-powered football analysis system that employs the YOLO (You Only Look Once) detection framework to achieve real-time identification and tracking of players, balls, and referees in football match videos. The system integrates computer vision techniques, including team classification through K-means clustering, optical flow for camera motion compensation, and homography transformation for perspective correction. Our implementation achieved remarkable performance metrics with 82.2% mAP50, 90.2% precision, and 77.0% recall across all object classes. The system successfully processes raw match footage to generate automated analytics, including player tracking, speed calculation, distance metrics, and possession statistics, offering an economical substitute for expensive GPS-based tracking systems used in professional sports.

Published by: Aryan Lalwani

Author: Aryan Lalwani

Paper ID: V11I5-1138

Paper Status: published

Published: September 15, 2025

Full Details
Review Paper

Crypto & Traditional Banking: An Evolving Relationship

The rapid rise of cryptocurrencies over the past decade has catalyzed a major shift in the global financial landscape. Once niche digital experiments, cryptocurrencies powered by blockchain and tokenization technologies have now achieved wider recognition across institutional investors, fintech innovators, and even central banks. A 2025 survey by EY and Coinbase revealed that 83% of institutional investors plan to increase their digital asset allocations, driven by growing regulatory clarity and product innovation.

Published by: Nitant Kalia

Author: Nitant Kalia

Paper ID: V11I5-1142

Paper Status: published

Published: September 15, 2025

Full Details
Research Paper

A Comparative Study of Machine Learning Algorithms for Real-Time DDoS Detection in Cloud Environments

Cloud environments are scalable and cost-effective, but they are also highly susceptible to cyberattacks, such as Distributed Denial of Service (DDoS) attacks, which can exhaust resources and impact availability. To counter these threats, this research investigates supervised machine learning techniques for identifying such attacks in real-time based on the BCCC cPacket Cloud DDoS 2024 dataset. Following deep preprocessing and exploratory data analysis, a multi-class classifier was employed to differentiate between benign, suspicious, and attack traffic. Of the models to be compared, the Decision Tree Classifier achieved the highest mark with an accuracy rate of 96.8 percent, which indicates its ability to categorize the majority of cases of traffic accordingly. The other models, including K Nearest Neighbors, Ridge Regression, Logistic Regression, and Linear Support Vector Classifier, had lower levels of accuracy, with Decision Tree always delivering the best. The results confirm that the Decision Tree is the best and most efficient model for precise real-time identification of DDoS attacks in cloud systems.

Published by: Oduwunmi Odukoya, Mariam Adetoun Sanusi, Samuel Adenekan

Author: Oduwunmi Odukoya

Paper ID: V11I5-1146

Paper Status: published

Published: September 11, 2025

Full Details
Research Paper

Performance Overview of Light Electric Vehicle Powered by Battery in Assistance with Supercapacitor

The growing electrical vehicle (EV) market has been taking the attention of customers & manufacturers in the last couple of years. EVs with lithium-ion (Li-ion) batteries stand as an alternate to the conventional power source of fossil fuel engines. While sustaining the stand, the limitations of battery as a sole power source surface, like range, temperature, state of health, dynamic power requirement & addressing different duty cycles. In this regard, a hybrid power source with a supercapacitor has emerged as a solution. The combined power source is known as a hybrid energy storage system (HESS). In this report we will see the current progress in technology for the supercapacitor-based powertrain from the viewpoint of range, charge-discharge cycle, overall weight of power source, power split strategies & ultimate cost impact. The HESS has basically three configurations—passive, semi-passive & active. The efficiency of the system improves with the introduction of power electronics. Comparative analysis of passive, semi-passive, and active HESS architectures confirms the superiority of active systems, which enable intelligent power-split strategies through advanced power electronics. Although they involve higher initial costs, lifecycle evaluations highlight reduced battery replacement frequency and lower operating expenses. This study establishes actively managed HESS as a scalable and sustainable solution for three-wheeler electrification, combining enhanced performance and long-term durability.

Published by: Amit Aundhakar, Sateesh Patil, Shashank Gawade

Author: Amit Aundhakar

Paper ID: V11I5-1141

Paper Status: published

Published: September 8, 2025

Full Details
Research Paper

A Deep Dive into the Adversarial Training Process of Modern LLMs

Large language models can be trained to resist jailbreak attempts in order to ensure the model’s safety. This literature review analyzes common adversarial training techniques used in modern LLMs to understand and break down the process of adversarial training. The paper looks into 6 training techniques, ranging from PGD to Deployment Time Safety Layers and Model Editing. For each method, we look into how they work behind the scenes, their strengths, weaknesses, and real-world usage. We then synthesize the most effective manner of training models against adversarial inputs. This research revealed that the most effective training technique emphasizes a multilayered defense strategy. Future research can look at improving model-editing coverage, creating open-sourced LLM benchmarks to test against jailbreaks, and closing the divide between embedding-space and discrete input training.

Published by: Armaan Mahajan

Author: Armaan Mahajan

Paper ID: V11I5-1137

Paper Status: published

Published: September 4, 2025

Full Details
Request a Call
If someone in your research area is available then we will connect you both or our counsellor will get in touch with you.

    [honeypot honeypot-378]

    X
    Journal's Support Form
    For any query, please fill up the short form below. Try to explain your query in detail so that our counsellor can guide you. All fields are mandatory.

      X
       Enquiry Form
      Contact Board Member

        Member Name

        [honeypot honeypot-527]

        X
        Contact Editorial Board

          X

            [honeypot honeypot-310]

            X