A deep ensemble learning method for single finger-vein identification
Finger-vein biometrics has been extensively investigated for personal verification. Single sample per person (SSPP) finger-vein recognition is one of the open issues in finger-vein recognition. Despite recent advances in deep neural networks for finger-vein recognition, current approaches depend on...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2023-01-01
|
| Series: | Frontiers in Neurorobotics |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fnbot.2022.1065099/full |
| _version_ | 1797956163597238272 |
|---|---|
| author | Chongwen Liu Chongwen Liu Huafeng Qin Huafeng Qin Qun Song Qun Song Huyong Yan Huyong Yan Fen Luo Fen Luo |
| author_facet | Chongwen Liu Chongwen Liu Huafeng Qin Huafeng Qin Qun Song Qun Song Huyong Yan Huyong Yan Fen Luo Fen Luo |
| author_sort | Chongwen Liu |
| collection | DOAJ |
| description | Finger-vein biometrics has been extensively investigated for personal verification. Single sample per person (SSPP) finger-vein recognition is one of the open issues in finger-vein recognition. Despite recent advances in deep neural networks for finger-vein recognition, current approaches depend on a large number of training data. However, they lack the robustness of extracting robust and discriminative finger-vein features from a single training image sample. A deep ensemble learning method is proposed to solve the SSPP finger-vein recognition in this article. In the proposed method, multiple feature maps were generated from an input finger-vein image, based on various independent deep learning-based classifiers. A shared learning scheme is investigated among classifiers to improve their feature representation captivity. The learning speed of weak classifiers is also adjusted to achieve the simultaneously best performance. A deep learning model is proposed by an ensemble of all these adjusted classifiers. The proposed method is tested with two public finger vein databases. The result shows that the proposed approach has a distinct advantage over all the other tested popular solutions for the SSPP problem. |
| first_indexed | 2024-04-10T23:44:36Z |
| format | Article |
| id | doaj.art-4fb937e0bfdc45fea15a03532b30a27c |
| institution | Directory Open Access Journal |
| issn | 1662-5218 |
| language | English |
| last_indexed | 2024-04-10T23:44:36Z |
| publishDate | 2023-01-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Neurorobotics |
| spelling | doaj.art-4fb937e0bfdc45fea15a03532b30a27c2023-01-11T05:08:33ZengFrontiers Media S.A.Frontiers in Neurorobotics1662-52182023-01-011610.3389/fnbot.2022.10650991065099A deep ensemble learning method for single finger-vein identificationChongwen Liu0Chongwen Liu1Huafeng Qin2Huafeng Qin3Qun Song4Qun Song5Huyong Yan6Huyong Yan7Fen Luo8Fen Luo9College of Artificial Intelligence, Chongqing Technology and Business University, Chongqing, ChinaChongqing Key Laboratory of Intelligent Perception and BlockChain Technology, Chongqing Technology and Business University, Chongqing, ChinaCollege of Artificial Intelligence, Chongqing Technology and Business University, Chongqing, ChinaChongqing Key Laboratory of Intelligent Perception and BlockChain Technology, Chongqing Technology and Business University, Chongqing, ChinaCollege of Artificial Intelligence, Chongqing Technology and Business University, Chongqing, ChinaChongqing Key Laboratory of Intelligent Perception and BlockChain Technology, Chongqing Technology and Business University, Chongqing, ChinaCollege of Artificial Intelligence, Chongqing Technology and Business University, Chongqing, ChinaChongqing Key Laboratory of Intelligent Perception and BlockChain Technology, Chongqing Technology and Business University, Chongqing, ChinaCollege of Artificial Intelligence, Chongqing Technology and Business University, Chongqing, ChinaChongqing Key Laboratory of Intelligent Perception and BlockChain Technology, Chongqing Technology and Business University, Chongqing, ChinaFinger-vein biometrics has been extensively investigated for personal verification. Single sample per person (SSPP) finger-vein recognition is one of the open issues in finger-vein recognition. Despite recent advances in deep neural networks for finger-vein recognition, current approaches depend on a large number of training data. However, they lack the robustness of extracting robust and discriminative finger-vein features from a single training image sample. A deep ensemble learning method is proposed to solve the SSPP finger-vein recognition in this article. In the proposed method, multiple feature maps were generated from an input finger-vein image, based on various independent deep learning-based classifiers. A shared learning scheme is investigated among classifiers to improve their feature representation captivity. The learning speed of weak classifiers is also adjusted to achieve the simultaneously best performance. A deep learning model is proposed by an ensemble of all these adjusted classifiers. The proposed method is tested with two public finger vein databases. The result shows that the proposed approach has a distinct advantage over all the other tested popular solutions for the SSPP problem.https://www.frontiersin.org/articles/10.3389/fnbot.2022.1065099/fullfinger-vein recognitionsingle sample per persondeep learningensemble learningpattern recognition |
| spellingShingle | Chongwen Liu Chongwen Liu Huafeng Qin Huafeng Qin Qun Song Qun Song Huyong Yan Huyong Yan Fen Luo Fen Luo A deep ensemble learning method for single finger-vein identification Frontiers in Neurorobotics finger-vein recognition single sample per person deep learning ensemble learning pattern recognition |
| title | A deep ensemble learning method for single finger-vein identification |
| title_full | A deep ensemble learning method for single finger-vein identification |
| title_fullStr | A deep ensemble learning method for single finger-vein identification |
| title_full_unstemmed | A deep ensemble learning method for single finger-vein identification |
| title_short | A deep ensemble learning method for single finger-vein identification |
| title_sort | deep ensemble learning method for single finger vein identification |
| topic | finger-vein recognition single sample per person deep learning ensemble learning pattern recognition |
| url | https://www.frontiersin.org/articles/10.3389/fnbot.2022.1065099/full |
| work_keys_str_mv | AT chongwenliu adeepensemblelearningmethodforsinglefingerveinidentification AT chongwenliu adeepensemblelearningmethodforsinglefingerveinidentification AT huafengqin adeepensemblelearningmethodforsinglefingerveinidentification AT huafengqin adeepensemblelearningmethodforsinglefingerveinidentification AT qunsong adeepensemblelearningmethodforsinglefingerveinidentification AT qunsong adeepensemblelearningmethodforsinglefingerveinidentification AT huyongyan adeepensemblelearningmethodforsinglefingerveinidentification AT huyongyan adeepensemblelearningmethodforsinglefingerveinidentification AT fenluo adeepensemblelearningmethodforsinglefingerveinidentification AT fenluo adeepensemblelearningmethodforsinglefingerveinidentification AT chongwenliu deepensemblelearningmethodforsinglefingerveinidentification AT chongwenliu deepensemblelearningmethodforsinglefingerveinidentification AT huafengqin deepensemblelearningmethodforsinglefingerveinidentification AT huafengqin deepensemblelearningmethodforsinglefingerveinidentification AT qunsong deepensemblelearningmethodforsinglefingerveinidentification AT qunsong deepensemblelearningmethodforsinglefingerveinidentification AT huyongyan deepensemblelearningmethodforsinglefingerveinidentification AT huyongyan deepensemblelearningmethodforsinglefingerveinidentification AT fenluo deepensemblelearningmethodforsinglefingerveinidentification AT fenluo deepensemblelearningmethodforsinglefingerveinidentification |