Face Hallucination From New Perspective of Non-Linear Learning Compressed Sensing
The past decade has witnessed a prosperity of sparsity-inspired face hallucination methods that use sparse prior and instances to generate High-Resolution (HR) faces. However, they need numerous Low-Resolution (LR) and HR instance pairs and adopt approximate sparse coding, which will bring bias to t...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2020-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/8946533/ |
| _version_ | 1818920785055580160 |
|---|---|
| author | Shuyuan Yang Xiaoyang Hao Zhi Liu Chen Yang Min Wang |
| author_facet | Shuyuan Yang Xiaoyang Hao Zhi Liu Chen Yang Min Wang |
| author_sort | Shuyuan Yang |
| collection | DOAJ |
| description | The past decade has witnessed a prosperity of sparsity-inspired face hallucination methods that use sparse prior and instances to generate High-Resolution (HR) faces. However, they need numerous Low-Resolution (LR) and HR instance pairs and adopt approximate sparse coding, which will bring bias to the recovery and suffer from high computational burden. In this paper we advance a Single Face Image Hallucination (SFIH) method from a new perspective of Non-linear Learning Compressive Sensing (NLCS), which can recover HR faces from a surprisingly small number of HR faces. The nonlinear sparse coding of facial images is explored, and a Deep AutoEncoder (DAE) network is constructed for learning a kernel function from a single HR instance set. SFIH is then reduced to an analytic compressive recovery problem by reformulating linear sparse coding as a nonlinear DAE model. By exploring the nonlinear sparsity in the feature space, NLCS can accurately and rapidly recover HR facial images with large magnification factor and exhibit robustness to LR-HR instance pairs mapping. Some experiments are taken on realizing 3X, 6X, 9X amplification of face images, and the results prove its efficiency and superiority to its counterparts. |
| first_indexed | 2024-12-20T01:27:16Z |
| format | Article |
| id | doaj.art-636af86991d74e66be1ef33bac2e1097 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-12-20T01:27:16Z |
| publishDate | 2020-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-636af86991d74e66be1ef33bac2e10972022-12-21T19:58:12ZengIEEEIEEE Access2169-35362020-01-0189434944010.1109/ACCESS.2019.29633608946533Face Hallucination From New Perspective of Non-Linear Learning Compressed SensingShuyuan Yang0https://orcid.org/0000-0002-4796-5737Xiaoyang Hao1https://orcid.org/0000-0002-6062-841XZhi Liu2https://orcid.org/0000-0003-4781-7125Chen Yang3https://orcid.org/0000-0003-0034-2265Min Wang4https://orcid.org/0000-0002-7571-1662School of Artificial Intelligence, Xidian University, Xi’an, ChinaSchool of Artificial Intelligence, Xidian University, Xi’an, ChinaSchool of Artificial Intelligence, Xidian University, Xi’an, ChinaSchool of Artificial Intelligence, Xidian University, Xi’an, ChinaKey Laboratory of Radar Signal Processing, Xidian University, Xi’an, ChinaThe past decade has witnessed a prosperity of sparsity-inspired face hallucination methods that use sparse prior and instances to generate High-Resolution (HR) faces. However, they need numerous Low-Resolution (LR) and HR instance pairs and adopt approximate sparse coding, which will bring bias to the recovery and suffer from high computational burden. In this paper we advance a Single Face Image Hallucination (SFIH) method from a new perspective of Non-linear Learning Compressive Sensing (NLCS), which can recover HR faces from a surprisingly small number of HR faces. The nonlinear sparse coding of facial images is explored, and a Deep AutoEncoder (DAE) network is constructed for learning a kernel function from a single HR instance set. SFIH is then reduced to an analytic compressive recovery problem by reformulating linear sparse coding as a nonlinear DAE model. By exploring the nonlinear sparsity in the feature space, NLCS can accurately and rapidly recover HR facial images with large magnification factor and exhibit robustness to LR-HR instance pairs mapping. Some experiments are taken on realizing 3X, 6X, 9X amplification of face images, and the results prove its efficiency and superiority to its counterparts.https://ieeexplore.ieee.org/document/8946533/Face hallucinationnonlinear sparse codingnon-linear learning compressed sensingdeep autoencoder |
| spellingShingle | Shuyuan Yang Xiaoyang Hao Zhi Liu Chen Yang Min Wang Face Hallucination From New Perspective of Non-Linear Learning Compressed Sensing IEEE Access Face hallucination nonlinear sparse coding non-linear learning compressed sensing deep autoencoder |
| title | Face Hallucination From New Perspective of Non-Linear Learning Compressed Sensing |
| title_full | Face Hallucination From New Perspective of Non-Linear Learning Compressed Sensing |
| title_fullStr | Face Hallucination From New Perspective of Non-Linear Learning Compressed Sensing |
| title_full_unstemmed | Face Hallucination From New Perspective of Non-Linear Learning Compressed Sensing |
| title_short | Face Hallucination From New Perspective of Non-Linear Learning Compressed Sensing |
| title_sort | face hallucination from new perspective of non linear learning compressed sensing |
| topic | Face hallucination nonlinear sparse coding non-linear learning compressed sensing deep autoencoder |
| url | https://ieeexplore.ieee.org/document/8946533/ |
| work_keys_str_mv | AT shuyuanyang facehallucinationfromnewperspectiveofnonlinearlearningcompressedsensing AT xiaoyanghao facehallucinationfromnewperspectiveofnonlinearlearningcompressedsensing AT zhiliu facehallucinationfromnewperspectiveofnonlinearlearningcompressedsensing AT chenyang facehallucinationfromnewperspectiveofnonlinearlearningcompressedsensing AT minwang facehallucinationfromnewperspectiveofnonlinearlearningcompressedsensing |