ieee-publication

IEEE Published: Accurate Semantic Segmentation of Aerial Imagery Using Attention Res U-Net Architecture (IEEE ESCI 2024).

A project that analyzes historical Bitcoin prices to forecast future trends with deep learning and statistical models (LSTM, BiLSTM, ARIMA and Transformer). The Transformer-based approach achieves the highest accuracy, leveraging time-series data for improved financial decision-making.

Android computer vision app for real-time vegetable detection using smartphone camera input.

IEEE Published Research | Brain Tumor Classification using LBP Image Subtraction & KNN

IPL T20 score prediction using ensemble learning (XGBoost, Random Forest) + SVM, validated with EDA and KNN accuracy checks. Co-authored research published in IEEE ASSIC 2025

A video anomaly detection system using Markov Random Fields (MRFs) and autoencoders. It extracts frames, segments them using MRFs, and leverages autoencoders for anomaly detection. The system improves detection accuracy and reduces computation time by 60%, making it efficient for real-time surveillance applications.

Published IEEE Research (ICCCNT @ IIT Mandi): Lightweight on-device AST-based translation between dynamically typed interpreted languages (Python & JS).

Single-frame ego-trajectory prediction for autonomous vehicles. Published at IEEE IC2E3 2026 (Paper ID 835).

Published IEEE Research (AICAI 2024): Real-time detection and analysis of HTTP Command & Control (C2) botnets using MongoDB and machine learning.

Implementation of the IEEE IC2E3 2025 paper on optimal wearable sensor placement for human fall detection using machine learning models.

Implementation of the IEEE IATMSI 2026 paper on geriatric faller identification using foot-mounted inertial sensors and a hybrid 1D-CNN + Bi-LSTM architecture.

AI-powered Domain-Driven Design conformance for Python: turns SRS documents into a typed domain model, then detects architectural drift in VS Code via AST analysis and multi-agent LLM reasoning. Published at IISEC 2026 (IEEE Xplore).

An image-to-audio encryption framework with AI-enhanced data hiding for secure multimedia sharing. Using adversarial training and reinforcement learning, it ensures protection against unauthorized access while transforming visual data into immersive audio. It enhances security, creativity, and data integrity in multimedia applications.