Analyzing Beauty by Building Custom Profiles Using Machine Learning

Instead of tackling what is universal beauty, we propose to attempt to create profiles based on training data for individual understanding of beauty, while maintaining efficiency. This paper describes a fully functional software system called TBeauty that creates an individual machine-learning algor...

Full description

Saved in:
Bibliographic Details
Published in2019 IEEE 9th Annual Computing and Communication Workshop and Conference (CCWC) pp. 0372 - 0376
Main Authors Abrahams, Thomas, Bein, Doina
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.01.2019
Subjects
Classification algorithms
facial landmark point
facial mask
individualized beauty detection algorithm
Machine learning
Machine learning algorithms
Python
Software algorithms
software system
Online AccessGet full text

Cover

More Information
Summary:Instead of tackling what is universal beauty, we propose to attempt to create profiles based on training data for individual understanding of beauty, while maintaining efficiency. This paper describes a fully functional software system called TBeauty that creates an individual machine-learning algorithm based off a profile to allow users to get an accurate assessment of what they would classify and score beauty. The system would take data and use facial landmarks in order to assess this. The system can be changed to allow automatic search of new images to proactively select the ones that have high scores for any individual user. The project uses an existing 68-point landmark detection algorithm proposed by Rainer Lienhart to identify the facial landmark points (a so-called facial mask) from images that will be fed into various machine-learning algorithms. The software is available on GitHub.
DOI:10.1109/CCWC.2019.8666528