Media.Player

Method and System for Automatic Beauty Recognition and Fashion Recommendation

Technology Benefits

The system/method:ΓÇóAssesses the effect of cosmetics and clothingΓÇóAllows for a robust decision as compared to conventional approachesΓÇóInvolves the use of multiple layer auto-encoders for obtaining attractiveness-aware features, enabling human-like assessment of beautyΓÇóEffectively processes the given face or clothing images and assess the beauty from appearanceΓÇóAllows for an effective evaluation/analysis of makeup and fashion professionalism/skillsΓÇóAllows for development of an integration framework to rate makeup and/or fashion through predicted confidence scoringΓÇóIs applicable on a larger and more diverse dataset, avoiding potential restrictions on faces, poses, expressions and lighting conditions as observed with conventional mechanismsΓÇóWould be commercially useful for the following applications:oProfessional digital face makeup analysis/trainingoWeb/Mobile/Social/Fashion applications for entertainmentoCultural/Educational applications such as theatrical makeup recognitionoFilm making toolsoBiometrics systems for security applicationsoForensic science application toolsoOnline shopping recommendation and advertisement systemoPlastic surgery software

Detailed Technology Description

None

*Abstract

In prior art, the assessment/quantification of facial attractiveness is an ever-lasting issue. The facial beauty and its measurement has been widely debated for centuries. It has drawn a considerable attention even from social science and multimedia community recently. None of the existing techniques/mechanisms are capable of addressing this need effectively, as they are associated with one of more limitations, such as use of heavy duty landmark annotations and use of non-unified attractiveness criteria. This invention relates to a novel method and a system for automatic beauty assessment, makeup evaluation and fashion recommendation capable of overcoming the prior art limitations and fulfilling the unmet needs.

*Principal Investigator

Name: Yun Fu

Department:

Country/Region

USA