Journal of Applied Optics, Volume. 44, Issue 3, 565(2023)
Method for cross-age face recognition based on identity-age sharing features
Figures & Tables(8)
Table 1. Adopted data sets
View tableView in ArticleTable 1. Adopted data sets
Datasets Identity Images MS1MV2[ 3 ]85742 5774205 AgeDB-30[ 19 ]6000 pairs 12000 CACD-VS[ 20 ]4000 pairs 8000 CALFW[ 21 ]6000 pairs 12000 LFW[ 22 ]6000 pairs 12000
Table 2. Accuracy of different methods on four data sets
Acc/% View tableView in ArticleTable 2. Accuracy of different methods on four data sets
Acc/% Method AgeDB-30[ 19 ]CACD-VS[ 20 ]CALFW[ 21 ]LFW[ 22 ]DAL[ 9 ]95.98 99.40 — 99.47 MTLFace[ 10 ]96.23 99.55 95.62 99.52 ISF(Our) 97.07 99.58 95.73 99.73
Table 3. Ablation test
Acc/% View tableView in ArticleTable 3. Ablation test
Acc/% Model AgeDB-30 CACD-VS CALFW Baseline 96.23 99.14 95.27 Baseline + Age 96.25 99.18 95.30 ISF(Our) 97.07 99.58 95.73
Tools
Get Citation
Copy Citation Text
Zikang ZHANG, Songfeng YIN, Liangcai CAO, Cheng LIU. Method for cross-age face recognition based on identity-age sharing features[J]. Journal of Applied Optics, 2023, 44(3): 565
Paper Information
Category: Research Articles
Received: Aug. 22, 2022
Accepted: --
Published Online: Jun. 19, 2023
The Author Email: YIN Songfeng (yinsongfeng@tsinghua-hf.edu.cn)
© Copyright 2018-2021 | Chinese Laser Press.
All Rights Reserved 沪ICP备15018463号-20