Research On Application Design Of Makeup Recommendation For Chinese Women

According to the survey,appearance is the most important factor affecting Chinese women's self-confidence,but nearly 80% of Chinese women's makeup groups are still in the primary stage.For makeup beginners,they need to spend a lot of time exploring their ideal makeup;at the same time,a wealth of application systems based on face recognition have been produced in the beauty industry,but these application systems are still difficult to meet the demands of Chinese women for makeup recommendations.Through the user research,it is found that 84.2% of the Chinese female respondents have the demand for makeup recommendations,and their choices of makeup styles are adaptable to occupations and occasions,among which the demand for daily makeup is the greatest.To summarize,the target population,user scenarios and product positioning of this study are finally determined.Basing on certain theoretical knowledge and practical skills of makeup,we collected a face style classification data set and two makeup recommendation data sets.After manual classifications,professional make-up artists were invited to correct the face and makeup style classifications.In the matching experiment between the face and makeup styles,in order to evaluate the impact between different face and makeup styles,we randomly selected 15 pictures from each face style of the face style classification data set,and used a makeup migration algorithm to transfer different styles of makeup to the selected pictures,The results showed that there are significant differences in the effects of the same makeup on different styles of faces and the same face on different styles of makeup,and there is a positive relationship between faces with different feeling and perception levels and the idealized form and color mapping of makeup.To summarize,the recommendation rules and design principles of Chinese women's makeup are finally put forward.According to the results of the user research and the above experiment,we proposed a method of makeup recommendation for Chinese women based on the analysis of facial features.The classification algorithm based on neural network is used to supervise and train the classified data set,through which the face style classification model is obtained.According to certain makeup recommendation rules,the trained network branch of key point detection of faces is used to extract the feature maps of the input images and the corresponding images in the makeup recommendation data set,and the cosine similarity of the feature maps is compared and sorted in hidden space to select the most similar makeup images that are recommended to users.According to the results of the user research,makeup recommendation rules,the experiment between the face and makeup styles and the makeup recommendation algorithm,we designed a makeup recommendation application for Chinese women based on a We Chat applet to meet the users' needs,which can help users quickly find their favorite makeup though the Internet.In addition,we used the developed application prototype to test the system's usability,and the test results prove its great value on practice.