Year: 2022
Technologies: LAMP Stack (Linux, Apache, MySQL, PHP), C#, AWS Rekognition, Computer Vision
The Smart Wardrobe originated from a real-world usability problem and evolved into a full-stack computer vision project focused on sustainability, automation, and user-centred design. Inspired by interactive retail carousels such as McDonald’s touchscreen kiosks, Alyzande designed a digital wardrobe interface to visualise clothing, track usage, and reduce overbuying caused by clutter and poor visibility.
The system was built as a LAMP-stack CRUD web application with a MySQL database and a responsive carousel-based front end. Users could log garments, calculate cost per wear, and view sustainability metrics. Alyzande extended functionality through AWS Rekognition, applying image classification to automatically identify and tag clothing items from a webcam feed, transforming the wardrobe into a self-updating, AI-powered inventory.
This project demonstrates Alyzande’s pragmatic innovation and data-driven problem solving, translating everyday frustration into a scalable, human-centred product. It also contributed to her MSc research, benchmarking Rekognition’s performance against state-of-the-art clothing recognition models and confirming its value for applied AI in consumer contexts.
Built using the LAMP stack (Linux, Apache, MySQL, PHP) with a C#-driven front end. Integrated AWS Rekognition for real-time image classification and tagging. The application stored metadata such as brand, colour, and last-worn date to enable analytics and environmental insights. The code structure prioritised scalability, modularity, and the ability to integrate future IoT sensors or wardrobe-mounted tablets.