Face illumination processing via dense feature maps and multiple receptive fields
Abstract Recently, illumination processing of facial image based on generative adversarial networks has made favourable progress. However, the image quality is not so satisfactory and the recognition accuracy is low when the face image under extreme illumination conditions. For these reasons, an ela...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Wiley
2021-08-01
|
Series: | Electronics Letters |
Subjects: | |
Online Access: | https://doi.org/10.1049/ell2.12181 |
_version_ | 1811183592645066752 |
---|---|
author | Shenggui Ling Keren Fu Ye Lin Di You Peng Cheng |
author_facet | Shenggui Ling Keren Fu Ye Lin Di You Peng Cheng |
author_sort | Shenggui Ling |
collection | DOAJ |
description | Abstract Recently, illumination processing of facial image based on generative adversarial networks has made favourable progress. However, the image quality is not so satisfactory and the recognition accuracy is low when the face image under extreme illumination conditions. For these reasons, an elaborately‐designed architecture based on convolutional neural network and generative adversarial networks for processing face illumination is presented. A novel dense feature maps loss that computes loss by using the varisized feature maps extracted from different convolutional layers of pre‐trained feature network is put forward. Moreover, multiple‐receptive‐fields‐based generator that uses multiple encoders during encoding phase is also proposed, and these encoders have the same structure with different kernel size. A variety of experimental results demonstrate that the method is superior to the state‐of‐the‐art methods under various illumination challenges. Code will be available soon at https://github.com/ling20cn/IP‐GAN |
first_indexed | 2024-04-11T09:49:31Z |
format | Article |
id | doaj.art-3f6b9ea1e2a7495e8042aba7222b130d |
institution | Directory Open Access Journal |
issn | 0013-5194 1350-911X |
language | English |
last_indexed | 2024-04-11T09:49:31Z |
publishDate | 2021-08-01 |
publisher | Wiley |
record_format | Article |
series | Electronics Letters |
spelling | doaj.art-3f6b9ea1e2a7495e8042aba7222b130d2022-12-22T04:30:51ZengWileyElectronics Letters0013-51941350-911X2021-08-01571662762910.1049/ell2.12181Face illumination processing via dense feature maps and multiple receptive fieldsShenggui Ling0Keren Fu1Ye Lin2Di You3Peng Cheng4The National Key Laboratory of Fundamental Science on Synthetic Vision Sichuan University Chengdu Sichuan ChinaThe College of Computer Science Sichuan University Chengdu Sichuan ChinaThe National Key Laboratory of Fundamental Science on Synthetic Vision Sichuan University Chengdu Sichuan ChinaThe National Key Laboratory of Fundamental Science on Synthetic Vision Sichuan University Chengdu Sichuan ChinaThe School of Aeronautics and Astronautics Sichuan University Chengdu Sichuan ChinaAbstract Recently, illumination processing of facial image based on generative adversarial networks has made favourable progress. However, the image quality is not so satisfactory and the recognition accuracy is low when the face image under extreme illumination conditions. For these reasons, an elaborately‐designed architecture based on convolutional neural network and generative adversarial networks for processing face illumination is presented. A novel dense feature maps loss that computes loss by using the varisized feature maps extracted from different convolutional layers of pre‐trained feature network is put forward. Moreover, multiple‐receptive‐fields‐based generator that uses multiple encoders during encoding phase is also proposed, and these encoders have the same structure with different kernel size. A variety of experimental results demonstrate that the method is superior to the state‐of‐the‐art methods under various illumination challenges. Code will be available soon at https://github.com/ling20cn/IP‐GANhttps://doi.org/10.1049/ell2.12181Image and video codingImage recognitionComputer vision and image processing techniquesNeural nets |
spellingShingle | Shenggui Ling Keren Fu Ye Lin Di You Peng Cheng Face illumination processing via dense feature maps and multiple receptive fields Electronics Letters Image and video coding Image recognition Computer vision and image processing techniques Neural nets |
title | Face illumination processing via dense feature maps and multiple receptive fields |
title_full | Face illumination processing via dense feature maps and multiple receptive fields |
title_fullStr | Face illumination processing via dense feature maps and multiple receptive fields |
title_full_unstemmed | Face illumination processing via dense feature maps and multiple receptive fields |
title_short | Face illumination processing via dense feature maps and multiple receptive fields |
title_sort | face illumination processing via dense feature maps and multiple receptive fields |
topic | Image and video coding Image recognition Computer vision and image processing techniques Neural nets |
url | https://doi.org/10.1049/ell2.12181 |
work_keys_str_mv | AT shengguiling faceilluminationprocessingviadensefeaturemapsandmultiplereceptivefields AT kerenfu faceilluminationprocessingviadensefeaturemapsandmultiplereceptivefields AT yelin faceilluminationprocessingviadensefeaturemapsandmultiplereceptivefields AT diyou faceilluminationprocessingviadensefeaturemapsandmultiplereceptivefields AT pengcheng faceilluminationprocessingviadensefeaturemapsandmultiplereceptivefields |