Floor-ladder framework for human face beautification

  • Yulia Novskaya*
  • , Sun Ruoqi
  • , Hengliang Zhu
  • , Lizhuang Ma
  • *Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

In this paper, we propose a Floor-Ladder Framework (FLN) based on age evolution rules to generate beautified human faces. Beside the shape of faces, younger faces achieve more attractiveness. Thus we process the beautiful face by applying the reversed aging rules. Inspired by the layered optimization methods, the FLN adopts three floors and each floor contains two ladders: the Single Layer Older Neural Network (SLONN) and the extended Skull Model. The Peak Shift algorithm is designed to train the SLONN aiming to capture the reversed aging rules of the face skin. Due to the growth rules of the face shape, we extended the Skull Model by adding Marquardt Mask. Given the input portrait, our algorithm effectively produces a beautified human face without losing personal features.

Original languageEnglish
Title of host publicationAnalysis of Images, Social Networks and Texts - 6th International Conference, AIST 2017, Revised Selected Papers
EditorsAndrey V. Savchenko, Dmitry I. Ignatov, Sergei O. Kuznetsov, Irina A. Lomazova, Victor Lempitsky, Michael Khachay, Natalia Loukachevitch, Amedeo Napoli, Wil M. van der Aalst, Alexander Panchenko, Panos M. Pardalos, Stanley Wasserman
PublisherSpringer Verlag
Pages255-266
Number of pages12
ISBN (Print)9783319730127
DOIs
StatePublished - 2018
Externally publishedYes
Event6th International Conference on Analysis of Images, Social Networks and Texts, AIST 2017 - Moscow, Russian Federation
Duration: 27 Jul 201729 Jul 2017

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10716 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference6th International Conference on Analysis of Images, Social Networks and Texts, AIST 2017
Country/TerritoryRussian Federation
CityMoscow
Period27/07/1729/07/17

Keywords

  • Face beautification
  • Floor-ladder framework
  • The extended skull model
  • The peak shift algorithm
  • The single layer older neural network

Fingerprint

Dive into the research topics of 'Floor-ladder framework for human face beautification'. Together they form a unique fingerprint.

Cite this