Facial Age Estimation Models for Embedded Systems: A Comparative Study
Automated age estimation from face images is the process of assigning either an exact age or a specific age range to a facial image. In this paper a comparative study of the current techniques suitable for this task is performed, with an emphasis on lightweight models suitable for embedded implement...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2023-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10041926/ |
| _version_ | 1797905473658159104 |
|---|---|
| author | Zorana Dozdor Tomislav Hrkac Karla Brkic Zoran Kalafatic |
| author_facet | Zorana Dozdor Tomislav Hrkac Karla Brkic Zoran Kalafatic |
| author_sort | Zorana Dozdor |
| collection | DOAJ |
| description | Automated age estimation from face images is the process of assigning either an exact age or a specific age range to a facial image. In this paper a comparative study of the current techniques suitable for this task is performed, with an emphasis on lightweight models suitable for embedded implementation. We investigate both the suitable modern deep learning architectures for feature extraction and the variants of framing the problem itself as either classification, regression or soft label classification. The models are evaluated on Audience dataset for age group classification and FG-NET dataset for exact age estimation. To gather in-depth insights into automated age estimation and in contrast to existing studies, we additionally compare the performance of both classification and regression on the same dataset. We propose a novel loss function that combines regression and classification approaches and show that it outperforms other considered approaches. At the same time, with a lightweight backbone, such an architecture is suitable for implementation on embedded devices. |
| first_indexed | 2024-04-10T10:05:53Z |
| format | Article |
| id | doaj.art-96d9133e618143bf89750988f5e6250d |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-04-10T10:05:53Z |
| publishDate | 2023-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-96d9133e618143bf89750988f5e6250d2023-02-16T00:00:35ZengIEEEIEEE Access2169-35362023-01-0111142821429210.1109/ACCESS.2023.324405910041926Facial Age Estimation Models for Embedded Systems: A Comparative StudyZorana Dozdor0https://orcid.org/0000-0002-9194-4589Tomislav Hrkac1https://orcid.org/0000-0002-4362-2489Karla Brkic2https://orcid.org/0000-0003-1355-4398Zoran Kalafatic3https://orcid.org/0000-0001-8918-9070Faculty of Electrical Engineering and Computing, University of Zagreb, Zagreb, CroatiaFaculty of Electrical Engineering and Computing, University of Zagreb, Zagreb, CroatiaFaculty of Electrical Engineering and Computing, University of Zagreb, Zagreb, CroatiaFaculty of Electrical Engineering and Computing, University of Zagreb, Zagreb, CroatiaAutomated age estimation from face images is the process of assigning either an exact age or a specific age range to a facial image. In this paper a comparative study of the current techniques suitable for this task is performed, with an emphasis on lightweight models suitable for embedded implementation. We investigate both the suitable modern deep learning architectures for feature extraction and the variants of framing the problem itself as either classification, regression or soft label classification. The models are evaluated on Audience dataset for age group classification and FG-NET dataset for exact age estimation. To gather in-depth insights into automated age estimation and in contrast to existing studies, we additionally compare the performance of both classification and regression on the same dataset. We propose a novel loss function that combines regression and classification approaches and show that it outperforms other considered approaches. At the same time, with a lightweight backbone, such an architecture is suitable for implementation on embedded devices.https://ieeexplore.ieee.org/document/10041926/Age estimationcomputer visiondeep learningface detection |
| spellingShingle | Zorana Dozdor Tomislav Hrkac Karla Brkic Zoran Kalafatic Facial Age Estimation Models for Embedded Systems: A Comparative Study IEEE Access Age estimation computer vision deep learning face detection |
| title | Facial Age Estimation Models for Embedded Systems: A Comparative Study |
| title_full | Facial Age Estimation Models for Embedded Systems: A Comparative Study |
| title_fullStr | Facial Age Estimation Models for Embedded Systems: A Comparative Study |
| title_full_unstemmed | Facial Age Estimation Models for Embedded Systems: A Comparative Study |
| title_short | Facial Age Estimation Models for Embedded Systems: A Comparative Study |
| title_sort | facial age estimation models for embedded systems a comparative study |
| topic | Age estimation computer vision deep learning face detection |
| url | https://ieeexplore.ieee.org/document/10041926/ |
| work_keys_str_mv | AT zoranadozdor facialageestimationmodelsforembeddedsystemsacomparativestudy AT tomislavhrkac facialageestimationmodelsforembeddedsystemsacomparativestudy AT karlabrkic facialageestimationmodelsforembeddedsystemsacomparativestudy AT zorankalafatic facialageestimationmodelsforembeddedsystemsacomparativestudy |