Feature Analysis And Application Based On Cloud Platform Of Facial Beauty Attractiveness

The research on facial beauty has been paid a lot of attention.Recent studies have demonstrated high cross cultural agreement in attractiveness rating of faces.In computational science,facial beauty is also a popular subject.Data-driven facial analysis based on geometric features and skin texture features has inspired many related studies in the fields of computer vision and machine learning;deep learning also works well in this subject.Although feature extraction for facial beauty analysis has been investigated extensively,little attention has been paid to data collection in this regard.A publicly available facial beauty dataset is expected to facilitate further research in this field.Considering the actually situation,the face with rotating and translating is more common.we proposed some models for this problem.The main work of this paper includes the following aspects:1)This paper proposes a benchmark dataset,namely the SCUT-FBP dataset,which can be used for different facial beauty analysis problems.We verify the ratings in terms of the rating distribution,standard deviation,consistency,and self-consistency.2)A facial attractiveness prediction model based on geometric features and skin texture features is proposed.We introduce combined-geometrical feature,2-dimensional Gabor texture features,Cortex-like Mechanisms features and wavelet features.Gabor texture features,CLM features and Wavelet features are invariant to rotation and translation.The effectiveness of these features for transformed face is verified in this novel.3)We proposed a facial beauty prediction cloud platform,names How Beauty.How Beautyis an application for facial beauty prediction.It's academic and interesting,which canpromote the develop of our a facial beauty prediction model.