Abdullah.
Hello, I'm Abdullah, a passionate and dedicated computer science enthusiast with a keen interest in exploring the limitless possibilities of technology. I have a solid foundation in computer science principles and a commitment to continuous learning and innovation.
Developed a responsive and user-friendly login system with secure authentication and strong form validation to enhance user security. Designed a dynamic weather application that fetches real-time weather data using APIs, allowing users to search for weather conditions in any city effortlessly. The project emphasizes modern UI design, API integration, and a seamless user experience across all devices.
PhysioGo is an AI-driven physiotherapy app that helps patients perform exercises correctly using real-time pose estimation. It tracks repetitions and sets, provides instant feedback, and allows physiotherapists to assign exercises and monitor progress remotely. Built with Flutter and Python, the app integrates WebSockets for smooth live video processing, ensuring an engaging and efficient rehabilitation experience across mobile devices.
This project automates student clustering and seating arrangements for exams or classroom assignments using AI-based K-Means clustering. It categorizes students based on their domain (department) and assigns them to rooms while balancing faculty distribution. The system then generates CSV reports and automatically converts them into PDFs for easy sharing.
Developed an NLP-based text classification and translation system using Naïve Bayes, RNN, and LSTM. The classification model categorizes text using Naïve Bayes, while the deep learning-based RNN and LSTM models handle English-to-Urdu translation. Implemented in Python with Scikit-learn, TensorFlow, and Keras, the project includes text preprocessing, feature extraction, and model evaluation. This work highlights my expertise in machine learning, deep learning, and NLP applications.
A digital image processing project focused on pedestrian gender classification under class imbalance. The pipeline includes image augmentation and enhancement, feature extraction using HOG, LBP, GLCM, and deep features from VGG19. Features are fused and reduced via PCA, followed by classification using a Linear SVM. Evaluation is done using 10-fold cross-validation with accuracy, precision, recall, F1-score, and confusion matrix.
This project analyzes Copenhagen’s cycling market using real-world data from government sources, directories, and marketplaces. It explores trends, consumer behavior, and sustainability impacts through data visualization and predictive models like Linear Regression, Random Forest, SVM, and Facebook Prophet—highlighting the role of cycling in urban sustainability.
Kohinoor Grammar School
Punjab Group of Colleges
FAST National University of Computer and Emerging Sciences