Data augmentation with occluded facial features for age and gender estimation
Abstract Here, the feature occlusion, a data augmentation method that simulates real‐life challenges on the main features of the human face for age and gender recognition is proposed. Previous methods achieved promising results on constrained data sets with strict environmental settings, but the res...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2021-11-01
|
| Series: | IET Biometrics |
| Subjects: | |
| Online Access: | https://doi.org/10.1049/bme2.12030 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832546760750268416 |
|---|---|
| author | Lu En Lin Chang Hong Lin |
| author_facet | Lu En Lin Chang Hong Lin |
| author_sort | Lu En Lin |
| collection | DOAJ |
| description | Abstract Here, the feature occlusion, a data augmentation method that simulates real‐life challenges on the main features of the human face for age and gender recognition is proposed. Previous methods achieved promising results on constrained data sets with strict environmental settings, but the results on unconstrained data sets are still far from perfect. The proposed method adopted three simple occlusion techniques, blackout, random brightness, and blur, and each simulates a different kind of challenge that would be encountered in real‐world applications. A modified cross‐entropy loss that gives less penalty to the age predictions that land on the adjacent classes of the ground truth class is also proposed. The effectiveness of our proposed method is verified by implementing the augmentation method and modified cross‐entropy loss on two different convolution neural networks, the slightly modified AdienceNet and the slightly modified VGG16, to perform age and gender classification. The proposed augmentation system improves the age and gender classification accuracy of the slightly modified AdienceNet network by 6.62% and 6.53% on the Adience data set, respectively. The proposed augmentation system also improves the age and gender classification accuracy of the slightly modified VGG16 network by 6.20% and 6.31% on the Adience data set, respectively. |
| format | Article |
| id | doaj-art-bc4fb56ab8d4422493a127f20db52f2e |
| institution | Kabale University |
| issn | 2047-4938 2047-4946 |
| language | English |
| publishDate | 2021-11-01 |
| publisher | Wiley |
| record_format | Article |
| series | IET Biometrics |
| spelling | doaj-art-bc4fb56ab8d4422493a127f20db52f2e2025-02-03T06:47:18ZengWileyIET Biometrics2047-49382047-49462021-11-0110664065310.1049/bme2.12030Data augmentation with occluded facial features for age and gender estimationLu En Lin0Chang Hong Lin1Department of Electronic and Computer Engineering National Taiwan University of Science and Technology Taipei Taiwan ChinaDepartment of Electronic and Computer Engineering National Taiwan University of Science and Technology Taipei Taiwan ChinaAbstract Here, the feature occlusion, a data augmentation method that simulates real‐life challenges on the main features of the human face for age and gender recognition is proposed. Previous methods achieved promising results on constrained data sets with strict environmental settings, but the results on unconstrained data sets are still far from perfect. The proposed method adopted three simple occlusion techniques, blackout, random brightness, and blur, and each simulates a different kind of challenge that would be encountered in real‐world applications. A modified cross‐entropy loss that gives less penalty to the age predictions that land on the adjacent classes of the ground truth class is also proposed. The effectiveness of our proposed method is verified by implementing the augmentation method and modified cross‐entropy loss on two different convolution neural networks, the slightly modified AdienceNet and the slightly modified VGG16, to perform age and gender classification. The proposed augmentation system improves the age and gender classification accuracy of the slightly modified AdienceNet network by 6.62% and 6.53% on the Adience data set, respectively. The proposed augmentation system also improves the age and gender classification accuracy of the slightly modified VGG16 network by 6.20% and 6.31% on the Adience data set, respectively.https://doi.org/10.1049/bme2.12030entropyface recognitionfeature extractionimage classificationneural nets |
| spellingShingle | Lu En Lin Chang Hong Lin Data augmentation with occluded facial features for age and gender estimation IET Biometrics entropy face recognition feature extraction image classification neural nets |
| title | Data augmentation with occluded facial features for age and gender estimation |
| title_full | Data augmentation with occluded facial features for age and gender estimation |
| title_fullStr | Data augmentation with occluded facial features for age and gender estimation |
| title_full_unstemmed | Data augmentation with occluded facial features for age and gender estimation |
| title_short | Data augmentation with occluded facial features for age and gender estimation |
| title_sort | data augmentation with occluded facial features for age and gender estimation |
| topic | entropy face recognition feature extraction image classification neural nets |
| url | https://doi.org/10.1049/bme2.12030 |
| work_keys_str_mv | AT luenlin dataaugmentationwithoccludedfacialfeaturesforageandgenderestimation AT changhonglin dataaugmentationwithoccludedfacialfeaturesforageandgenderestimation |