Lipstick and laughs: Facial make-up recommendation system

The increasing demand for personalised beauty solutions has driven the development of innovative tools to assist users in selecting make-up products tailored to their individual needs. This project presents a web application designed to deliver personalised foundation and lipstick recommendations by integrating computer vision and image processing techniques. The application captures a user’s photo through a live camera feed, utilising the browser’s Media Devices API, under controlled conditions to ensure optimal brightness and face area. Leveraging OpenCV.js and face-api.js, the system analyses the captured image to detect skin tone and undertone, converting it to LAB colour space for accurate colour data extraction from specific facial regions. Pre-defined datasets containing foundation and lipstick shades including shade names, product lines, LAB colour values, undertones, finishes, and coverage are used to match the detected skin characteristics with suitable products. The intuitive interface provides real-time feedback, guiding users seamlessly through the process and presenting tailored recommendations. This project showcases the potential of beauty technology to offer accessible, personalised make-up solutions, enhancing user experience and satisfaction in the cosmetics selection process.