TY - GEN
T1 - Floor-ladder framework for human face beautification
AU - Novskaya, Yulia
AU - Ruoqi, Sun
AU - Zhu, Hengliang
AU - Ma, Lizhuang
N1 - Publisher Copyright:
© Springer International Publishing AG 2018.
PY - 2018
Y1 - 2018
N2 - 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.
AB - 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.
KW - Face beautification
KW - Floor-ladder framework
KW - The extended skull model
KW - The peak shift algorithm
KW - The single layer older neural network
UR - https://www.scopus.com/pages/publications/85039460665
U2 - 10.1007/978-3-319-73013-4_24
DO - 10.1007/978-3-319-73013-4_24
M3 - 会议稿件
AN - SCOPUS:85039460665
SN - 9783319730127
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 255
EP - 266
BT - Analysis of Images, Social Networks and Texts - 6th International Conference, AIST 2017, Revised Selected Papers
A2 - Savchenko, Andrey V.
A2 - Ignatov, Dmitry I.
A2 - Kuznetsov, Sergei O.
A2 - Lomazova, Irina A.
A2 - Lempitsky, Victor
A2 - Khachay, Michael
A2 - Loukachevitch, Natalia
A2 - Napoli, Amedeo
A2 - van der Aalst, Wil M.
A2 - Panchenko, Alexander
A2 - Pardalos, Panos M.
A2 - Wasserman, Stanley
PB - Springer Verlag
T2 - 6th International Conference on Analysis of Images, Social Networks and Texts, AIST 2017
Y2 - 27 July 2017 through 29 July 2017
ER -