DeepEthnic: Multi-label ethnic classification from face images

Katia Huri, Eli Omid David, Nathan S. Netanyahu

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

6 Scopus citations

Abstract

Ethnic group classification is a well-researched problem, which has been pursued mainly during the past two decades via traditional approaches of image processing and machine learning. In this paper, we propose a method of classifying an image face into an ethnic group by applying transfer learning from a previously trained classification network for large-scale data recognition. Our proposed method yields state-of-the-art success rates of 99.02%, 99.76%, 99.2%, and 96.7%, respectively, for the four ethnic groups: African, Asian, Caucasian, and Indian.

Original languageEnglish
Title of host publicationArtificial Neural Networks and Machine Learning – ICANN 2018 - 27th International Conference on Artificial Neural Networks, 2018, Proceedings
EditorsYannis Manolopoulos, Barbara Hammer, Vera Kurkova, Lazaros Iliadis, Ilias Maglogiannis
PublisherSpringer Verlag
Pages604-612
Number of pages9
ISBN (Print)9783030014230
DOIs
StatePublished - 2018
Event27th International Conference on Artificial Neural Networks, ICANN 2018 - Rhodes, Greece
Duration: 4 Oct 20187 Oct 2018

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume11141 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference27th International Conference on Artificial Neural Networks, ICANN 2018
Country/TerritoryGreece
CityRhodes
Period4/10/187/10/18

Bibliographical note

Publisher Copyright:
© Springer Nature Switzerland AG 2018.

Funding

As previously indicated, the purpose of this research is to distinguish between the four ethnic groups: African. Asian, Caucasian, and Indian. We created our dataset by combining 10 different databases, originally proposed for the problem of face recognition, and then sorting them into the ethnic groups of interest. The databases included IMFDB [16], CNBC [20], Labeled Faces in the Wild (LFW) [7], the Essex face dataset [18], Face Tracer [10], the Yale face database [3], SCUT5000 [24], and additional collected image datasets. We also used the well-known FERET database [14,15], which contains facial images collected under the FERET program, sponsored at the time by the U.S. Department of Defense (DoD). Altogether, the collected dataset contains images of various sizes.

FundersFunder number
U.S. Department of DefenseDoD
Directorate for Social, Behavioral and Economic Sciences0339122

    Fingerprint

    Dive into the research topics of 'DeepEthnic: Multi-label ethnic classification from face images'. Together they form a unique fingerprint.

    Cite this