Title :

Deep learning based face beauty prediction via dynamic robust losses and ensemble regression

Author(s) :

Bougourzi, F. [Auteur]
Dornaika, F. [Auteur]
Ikerbasque - Basque Foundation for Science
Henan Polytechnic University
Universidad del Pais Vasco / Euskal Herriko Unibertsitatea [Espagne] [UPV/EHU]
Taleb-Ahmed, Abdelmalik [Auteur]
COMmunications NUMériques - IEMN [COMNUM - IEMN]
Institut d’Électronique, de Microélectronique et de Nanotechnologie - UMR 8520 [IEMN]

Journal title :

Knowledge-Based Systems

Pages :

108246

Publisher :

Elsevier

Publication date :

2022-04

ISSN :

0950-7051

English keyword(s) :

Facial beauty prediction
Convolutional neural network
Deep learning
Ensemble regression
Robust loss functions

HAL domain(s) :

Sciences de l'ingénieur [physics]

English abstract : [en]

In the last decade, several studies have shown that facial attractiveness can be learned by machines. In this paper, we address Facial Beauty Prediction from static images. The paper contains three main contributions. ...
Show more >In the last decade, several studies have shown that facial attractiveness can be learned by machines. In this paper, we address Facial Beauty Prediction from static images. The paper contains three main contributions. First, we propose a two-branch architecture (REX-INCEP) based on merging the architecture of two already trained networks to deal with the complicated high-level features associated with the FBP problem. Second, we introduce the use of a dynamic law to control the behaviour of the following robust loss functions during training: ParamSmoothL1, Huber and Tukey. Third, we propose an ensemble regression based on Convolutional Neural Networks (CNNs). In this ensemble, we use both the basic networks and our proposed network (REX-INCEP). The proposed individual CNN regressors are trained with different loss functions, namely MSE, dynamic ParamSmoothL1, dynamic Huber and dynamic Tukey. Our approach is evaluated on the SCUT-FBP5500 database using the two evaluation scenarios provided by the database creators: 60%–40% split and five-fold cross-validation. In both evaluation scenarios, our approach outperforms the state of the art on several metrics. These comparisons highlight the effectiveness of the proposed solutions for FBP. They also show that the proposed dynamic robust losses lead to more flexible and accurate estimators.Show less >

Language :

Anglais

Peer reviewed article :

Oui

Audience :

Internationale

Popular science :

Non

Collections :

• Institut d'Électronique, de Microélectronique et de Nanotechnologie (IEMN) - UMR 8520

Source :

Harvested from HAL