Skin Tone Detection and Foundation Recommendation Application Using CNN and C-BF
Year: 2026
Type: article
Abstract: The face is a crucial part of the body that requires attention in daily life, as it serves as one of the key attractions for an individual and can depict their personality. Having a flawless and beautiful face is a desire shared by many, especially women. Therefore, numerous women aspire to enhance their appearance through various means, one of which involves using facial cosmetics, such as foundation, to conceal imperfections. In research focusing on the application of Machine Learning (ML) and Deep Learning (DL) in the fields of Computer Vision (CV) and Data Science (DS), algorithms like Convolutional Neural Network (CNN) and Content-Based Filtering (C-BF) were employed. This quantitative study involved the collection and analysis of data to achieve optimal results, following the AI project cycle procedure for designing and implementing a facial skin color detection and foundation recommendation application. The application of the CNN algorithm yielded an accuracy of 96% with a sample size of 7200 images. Additionally, the implementation of C-BF provided 10 foundation recommendations with the highest similarity values among products based on facial skin color. The outcomes of these applications aim to assist users in finding a suitable foundation product that matches their skin color and characteristics. This not only saves time and effort in the foundation selection process but also enhances user satisfaction with the use of foundation products.
DOI: 10.47002/3f91kg44
Institution: Universitas Mulia
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W7225814856