Kao researcher explores machine learning for makeup evaluation with potential for personalisation
A Kao researcher has developed a machine learning model to objectively analyse and evaluate makeup texture on the skin with potential applications in the personalized cosmetics space. The research utilized deep neural networks (DNNs), a class of machine learning algorithms that aims to mimic the information processing of the brain. The model was able to accurately analyze and classify various skin attributes, such as age range, the use of base makeup, and makeup formulation type.
Traditionally, makeup evaluations were conducted by human experts. However, using DNNs could offer more accurate and objective assessments. The study concluded that deep learning technology has the potential to obtain a makeup finish evaluation technology that can evaluate subtle textures as well as human visual evaluation.
DNNs have previously been applied to skin evaluation, particularly in the field of medicine for detecting skin cancer. In this study, the DNN model was trained using skin patches – small sections of facial skin images. By training the model on these patches, it was able to retain fine texture, eliminate false correlations from non-skin features, and visualize the inferred results for the entire face.
The trained DNNs were able to classify various skin attributes, such as age range and presence of base makeup. It could also distinguish between different makeup formulations, such as liquid and powder. Additionally, the DNN model could quantify the perceived makeup feel and visualize its distribution across the face. This makeup feel refers to the sensation of makeup on the skin.
The research demonstrated that the model could accurately assess makeup conditions and provide insights into various skin texture aspects. The trained DNNs showed high prediction accuracy for the experts’ visual assessment. Moreover, the application of DNN to the evaluation of actual makeup conditions resulted in appropriate evaluation results in line with the appearance of the makeup finish.
The capabilities of this machine learning model make it a valuable tool for developing personalized cosmetics tailored to individual preferences and characteristics. The researchers concluded that further studies are needed to explore its potential in the personalized beauty space, including analyzing different skin conditions and developing customized cosmetics.
Overall, this research opens up new possibilities for the cosmetic industry to enhance makeup evaluation and provide more personalized beauty solutions. With machine learning technology, the accuracy and objectivity of makeup assessments can be improved, leading to a better understanding of makeup texture and formulation. Ultimately, this could revolutionize the way we approach cosmetics and empower individuals to find products that best suit their unique needs and preferences.
