VOLUME 3, ISSUE 6, JUNE 2026

Abstract: Data Structures and Algorithms (DSA) are the backbone of computer science education; still, traditional learning methods based on static code examples and slide-show lectures are ineffective in communicating the step-by- step dynamic process of algorithms. This drawback makes it difficult for students to form correct mental images of algorithmic processes, leading to passive learning and superficial understanding. Recent breakthroughs in Artificial Intelligence (AI), specifically Large Language Models (LLMs), open new avenues for developing intelligent and interactive learning environments that adapt to the needs of individual learners and allow multiple modalities of interaction.
This paper introduces the conceptualization and design of AlgoVista, an AI-augmented algorithm learning platform that combines real-time visualizations of algorithms with adaptive explanations and interactive assessments using LLMs. The proposed system combines a MERN-stack web interface with Python algorithm execution, an LLM for context- dependent explanations, and a text-to-speech system for audio narration. As algorithms run step by step, the system dynamically produces plain-language explanations for each state transition and provides short-form personalized quizzes based on the learner’s past interactions and mistakes. AlgoVista, by integrating deterministic algorithm execution, multimodal explanation, and learner-centric assessment in a unified framework, seeks to convert passive algorithm demonstrations into active learning experiences.The proposed architecture remedies the major shortcomings of existing visualization systems by allowing adaptability, interactivity, and continuous feedback, which help to facilitate conceptual understanding and learning outcomes in Data Structure and Algorithm education.

Keywords: LLM, Natural Language Process, Artificial Intelligence, Algorithm Visualization

Abstract: One of the main results of the validation system is based on the fingerprint-based iris recognition system and respective technology. The entire biometric process is very much authentic and unique than the other types of recognition system and validation process. This has provided innovative ideas in the daily lives of human beings. The multimodal biometric process has generally ap- plied various types of applications for properly dealing with the appropriate and most significant limitations of the “unimodal biometric system”. The entire process has been generally included with the proper sensitivity of noise, the population coverage areas, variability cases of the inter class and intra class issues, vulnerability cases of possible hacking and the non-universality criteria. The entire research paper has been mainly focused on the deep learning-oriented machine learning system. The fingerprint-based iris recognition system to do the proper validation of human beings has been mainly done by convolutional neural network (CNN) technique. In the existing data validation process, the iris recognition system has been mainly done with respect to the “high security protection system with actual fingerprints''. The entire paper has been briefly elaborated on the best uniqueness, reliability process and the proper “validity of the iris biometric vali- dation system” for the actual purpose of the person identification.

Keywords: Iris Recognition, Voter Authentication System, Artificial Intelligence in Voting, Identity Management, Fraud Prevention and Indian Election System.

Abstract: Epileptic seizures are sudden neurological events that require timely detection and intervention. Electroencephalogram (EEG) signals, due to their non-invasive nature, are widely used for seizure analysis. This paper presents a complete automated seizure detection framework using the Children’s Hospital Boston-Massachusetts Institute of Technology (CHB-MIT) Scalp EEG Database. The system integrates preprocessing, feature extraction, data balancing, data augmentation, and classification into a robust pipeline. Key features were derived using power spectral density, wavelet decomposition, entropy, and statistical measures. To overcome class imbalance, the Synthetic Minority Oversampling Technique (SMOTE) was employed. Multiple machine learning classifiers, including Support Vector Machine (SVM), Random Forest (RF), Extra Trees (ET), Logistic Regression (LR), XG Boost (XGB), and Light GBM (LGBM) were evaluated. Performance was assessed using 5-fold stratified cross-validation, demonstrating that the integration of spectral and nonlinear EEG features with advanced machine learning methods significantly improves seizure detection accuracy.

Keywords: Epileptic seizure detection; Electroencephalogram (EEG); Machine learning, Feature extraction; Children’s Hospital Boston- Massachusetts Institute of Technology (CHB-MIT).

Abstract: Deepfake technology, powered by advanced generative models like GANs and diffusion models, poses significant threats to media authenticity, privacy, and democratic processes by creating highly realistic manipulated videos and images. This research paper explores deepfake detection using Transformer architectures, particularly Vision Transformers (ViTs), which excel at capturing global contextual dependencies and subtle artifacts often missed by traditional CNNs. The study provides a comprehensive review of literature, focusing on contributions from Indian researchers, proposes a hybrid methodology integrating ViTs with spatiotemporal analysis, and evaluates its effectiveness.

Objectives include surveying state-of-the-art techniques, developing a robust detection model, analyzing performance on benchmark datasets like FaceForensics++, Celeb-DF, and DFDC, and discussing generalization challenges against evolving deepfakes. The proposed methodology employs a shallow or hybrid ViT backbone with attention mechanisms for efficient feature extraction from facial patches, combined with temporal modeling for video sequences. Experimental results demonstrate superior accuracy and efficiency compared to CNN baselines, achieving high detection rates while maintaining computational feasibility for real-time applications.

Key challenges addressed include cross-dataset generalization, robustness to compression and perturbations, and explainability. Discussion highlights the superiority of Transformers in modeling long-range dependencies and frequency-domain inconsistencies. The paper concludes with future directions, emphasizing multimodal approaches, adversarial training, and ethical deployment. This work contributes to the growing body of knowledge in digital forensics, advocating for collaborative efforts to combat misinformation. With deepfakes proliferating on social media, Transformer-based detectors offer a promising pathway toward trustworthy media ecosystems. (248 words)

Keywords: Deepfake Detection, Vision Transformers, ViT, Spatiotemporal Analysis, Digital Forensics, Generative AI, Cross-Dataset Generalization, Multimodal Detection

Abstract: Multi-Agent AI Systems (MAS) have emerged as a cornerstone of modern artificial intelligence, enabling multiple autonomous agents to interact, collaborate, negotiate, and coordinate in dynamic environments to achieve individual or shared objectives. This paradigm shifts from centralized, monolithic AI architectures to decentralized, emergent intelligence capable of tackling complex, real-world problems characterized by uncertainty, scalability demands, and distributed decision-making. This research paper provides a comprehensive examination of MAS, tracing their evolution from foundational Distributed Artificial Intelligence (DAI) concepts to contemporary integrations with Large Language Models (LLMs), reinforcement learning (RL), and edge computing.

The study emphasizes applications relevant to India's socio-economic context, including rural energy management, healthcare delivery, agriculture advisory, and enterprise automation. It synthesizes insights from Indian research contributions, identifies gaps, and proposes a hybrid MAS framework. Key challenges such as coordination overhead, security vulnerabilities, ethical alignment, and interoperability are critically analyzed alongside opportunities for scalable, resilient systems.

Findings from simulated experiments demonstrate that the proposed hybrid framework improves task completion efficiency by 28-42% compared to single-agent baselines in microgrid optimization and software debugging scenarios, with notable gains in adaptability under variable conditions. However, communication costs rise with agent count, underscoring the need for optimized protocols.

This paper advocates for responsible, context-aware deployment of MAS to support sustainable development goals. By fostering collective intelligence, MAS can address pressing national challenges while contributing to global AI advancements. (Word count: 248)

Keywords: Multi-Agent Systems, Agentic AI, Large Language Models, Collaborative Intelligence, Reinforcement Learning, Microgrids, Rural Healthcare, Ethical AI, Emergent Behavior, Hybrid Architectures