AI-driven Detection & Classification of Diabetic Retinopathy" is a cutting-edge approach that combines artificial intelligence (AI) with medical image analysis to identify and categorize a specific eye disease called Diabetic Retinopathy (DR).

No description available

The AI-Driven Threat Detection System is a comprehensive solution designed for military applications, leveraging advanced computer vision and machine learning techniques to detect potential threats in real-time.

Develop an AI-based system to detect and respond to cyber threats in real time. This project demonstrates skills in cybersecurity, machine learning, and anomaly detection.

AI-Driven UAV Collision & Intrusion Detection System enhances drone safety by combining YOLOv8 object detection, synthetic flight path generation, collision prediction, and intrusion alerts. A real-time Streamlit dashboard consolidates risks and events, ensuring safer UAV operations in congested airspaces.

The PlantPatrol project is an innovative initiative by a group of dedicated students from the Catholic University of Eastern Africa.

No description available

No description available

AI-driven phishing detection system using machine learning (BERT) for email classification and Bloom filters for URL validation. It identifies phishing emails by analyzing content, headers, and URLs, and checks domain legitimacy against a trusted dataset. Fast and efficient with Python and ML.

Final year project

An end-to-end cybersecurity platform that leverages machine learning and deep learning to detect, classify, and respond to network threats in real time. Supports NSL-KDD, CICIDS2017/2018, UNSW-NB15, and Bot-IoT datasets. Built with Python, FastAPI, Streamlit, PyTorch, LightGBM, Scikit-Learn, and Docker for reproducible deployment.

AI-Driven UAV Collision & Intrusion Detection System enhances drone safety by combining YOLOv8 object detection, synthetic flight path generation, collision prediction, and intrusion alerts. A real-time Streamlit dashboard consolidates risks and events, ensuring safer UAV operations in congested airspaces.

AI-driven Intrusion Detection System for IoT networks using XGBoost, trained on CICIoT2023 dataset. Achieves 99.25% accuracy on 34 attack types. Integrated with Suricata for real-time packet logging and Telegram for instant alerts. Built for secure, scalable IoT environments.

This project presents an AI-driven early warning system for sepsis detection in ICUs, using ML, DL, and temporal modeling. The system outperforms traditional methods with an ROC-AUC of 0.91 and an F1-score of 0.87. Future work focuses on multi-center validation and real-time integration.

AI-Driven Fake Hire Detection System for Identifying and Preventing Recruitment Scams

No description available

This project presents a Domestic AI-Driven Intrusion Detection System, an intelligent solution that safeguards home and small-scale networks using machine learning-based anomaly detection, threat intelligence, and a real-time dashboard for smart, autonomous, and explainable network defense.

A machine learning project focused on detecting fraudulent financial transactions using pattern analysis and predictive modeling.

This project focuses on developing an AI-powered fake news detection system using natural language processing and machine learning. It involves exploratory data analysis, text preprocessing, TF-IDF vectorization, and ensemble model optimization to classify news articles as real or fake, helping to combat misinformation in digital media.

No description available

No description available

Description: This project focuses on leveraging artificial intelligence and computer vision to improve agricultural efficiency by automating weed detection and herbicide spraying. The system integrates machine learning models trained on image datasets to accurately identify and classify weeds in crop fields.

No description available

No description available

No description available

AI-powered Defect Detection & Classification Solutions for Manufacturing Industries

SecIDS ist ein leistungsfähiges, KI-gestütztes Echtzeit-IDS (Intrusion Detection System)

No description available

AI-powered pipeline leak detection system using ML models (Random Forest/SVM) with 92% accuracy. Enables real-time alerts and predictive maintenance via SCADA integration.

This project presents an AI-powered embedded robotic system that automatically detects and removes weeds from agricultural fields using Convolutional Neural Networks (CNN), sensor fusion, and precision actuation. The system is designed to eliminate herbicide usage, protect crops, and automate weed management with high accuracy in real time.