Minutiae-Based Weighting Aggregation of Deep Convolutional Features for Vein Recognition
Deep convolutional neural network (DCNN) has achieved an outstanding performance in large-scale image recognition task because of its discriminative feature representation ability, and pre-trained DCNN models trained for one task have also been applied to domains that are different from their origin...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2018-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/8493468/ |
| _version_ | 1831645375391334400 |
|---|---|
| author | Jun Wang Kai Yang Zaiyu Pan Guoqing Wang Ming Li Yulian Li |
| author_facet | Jun Wang Kai Yang Zaiyu Pan Guoqing Wang Ming Li Yulian Li |
| author_sort | Jun Wang |
| collection | DOAJ |
| description | Deep convolutional neural network (DCNN) has achieved an outstanding performance in large-scale image recognition task because of its discriminative feature representation ability, and pre-trained DCNN models trained for one task have also been applied to domains that are different from their original purposes. Inspired by this idea, a novel hand-dorsa vein recognition model is constructed by adopting DCNN pre-trained on a large-scale database as a universal feature descriptor. However, due to the sparse distribution property of vein information, it is difficult to employ pre-trained DCNN model to extract discriminative deep convolutional features. Therefore, to obtain useful and discriminative deep convolutional features, a novel minutiae-based weighting aggregation (MWA) method is proposed. In specific, the proposed global max-pooling of preserving spatial position information is applied on the feature maps of convolutional layer to localize the minutiae of vein information, and then the minutiae feature of vein image is regarded as the mask that is named as minutiae feature mask, to select deep convolutional features that contain minutiae feature information of vein image. The final feature representation is formed by concatenating each selected deep convolutional feature that is generated by each minutiae feature mask. Series rigorous experiments on the lab-made database are conducted to evidence the effectiveness and feasibility of the proposed MWA for vein recognition. What's more, an additional experiment with subset of PolyU database illustrates its generalization ability and robustness. |
| first_indexed | 2024-12-19T13:34:40Z |
| format | Article |
| id | doaj.art-fec9df80039a4d1183773b83b6ca9c00 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-12-19T13:34:40Z |
| publishDate | 2018-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-fec9df80039a4d1183773b83b6ca9c002022-12-21T20:19:15ZengIEEEIEEE Access2169-35362018-01-016616406165010.1109/ACCESS.2018.28763968493468Minutiae-Based Weighting Aggregation of Deep Convolutional Features for Vein RecognitionJun Wang0Kai Yang1Zaiyu Pan2https://orcid.org/0000-0003-2946-6296Guoqing Wang3Ming Li4Yulian Li5School of Information and Control Engineering, China University of Mining and Technology, Xuzhou, ChinaSchool of Information and Control Engineering, China University of Mining and Technology, Xuzhou, ChinaSchool of Information and Control Engineering, China University of Mining and Technology, Xuzhou, ChinaSchool of Information and Control Engineering, China University of Mining and Technology, Xuzhou, ChinaSchool of Information and Control Engineering, China University of Mining and Technology, Xuzhou, ChinaSchool of Information and Control Engineering, China University of Mining and Technology, Xuzhou, ChinaDeep convolutional neural network (DCNN) has achieved an outstanding performance in large-scale image recognition task because of its discriminative feature representation ability, and pre-trained DCNN models trained for one task have also been applied to domains that are different from their original purposes. Inspired by this idea, a novel hand-dorsa vein recognition model is constructed by adopting DCNN pre-trained on a large-scale database as a universal feature descriptor. However, due to the sparse distribution property of vein information, it is difficult to employ pre-trained DCNN model to extract discriminative deep convolutional features. Therefore, to obtain useful and discriminative deep convolutional features, a novel minutiae-based weighting aggregation (MWA) method is proposed. In specific, the proposed global max-pooling of preserving spatial position information is applied on the feature maps of convolutional layer to localize the minutiae of vein information, and then the minutiae feature of vein image is regarded as the mask that is named as minutiae feature mask, to select deep convolutional features that contain minutiae feature information of vein image. The final feature representation is formed by concatenating each selected deep convolutional feature that is generated by each minutiae feature mask. Series rigorous experiments on the lab-made database are conducted to evidence the effectiveness and feasibility of the proposed MWA for vein recognition. What's more, an additional experiment with subset of PolyU database illustrates its generalization ability and robustness.https://ieeexplore.ieee.org/document/8493468/Pre-trained DCNNhand-dorsa vein recognitionminutiae-based weighting aggregationglobal max-pooling of preserving spatial position informationminutiae feature mask |
| spellingShingle | Jun Wang Kai Yang Zaiyu Pan Guoqing Wang Ming Li Yulian Li Minutiae-Based Weighting Aggregation of Deep Convolutional Features for Vein Recognition IEEE Access Pre-trained DCNN hand-dorsa vein recognition minutiae-based weighting aggregation global max-pooling of preserving spatial position information minutiae feature mask |
| title | Minutiae-Based Weighting Aggregation of Deep Convolutional Features for Vein Recognition |
| title_full | Minutiae-Based Weighting Aggregation of Deep Convolutional Features for Vein Recognition |
| title_fullStr | Minutiae-Based Weighting Aggregation of Deep Convolutional Features for Vein Recognition |
| title_full_unstemmed | Minutiae-Based Weighting Aggregation of Deep Convolutional Features for Vein Recognition |
| title_short | Minutiae-Based Weighting Aggregation of Deep Convolutional Features for Vein Recognition |
| title_sort | minutiae based weighting aggregation of deep convolutional features for vein recognition |
| topic | Pre-trained DCNN hand-dorsa vein recognition minutiae-based weighting aggregation global max-pooling of preserving spatial position information minutiae feature mask |
| url | https://ieeexplore.ieee.org/document/8493468/ |
| work_keys_str_mv | AT junwang minutiaebasedweightingaggregationofdeepconvolutionalfeaturesforveinrecognition AT kaiyang minutiaebasedweightingaggregationofdeepconvolutionalfeaturesforveinrecognition AT zaiyupan minutiaebasedweightingaggregationofdeepconvolutionalfeaturesforveinrecognition AT guoqingwang minutiaebasedweightingaggregationofdeepconvolutionalfeaturesforveinrecognition AT mingli minutiaebasedweightingaggregationofdeepconvolutionalfeaturesforveinrecognition AT yulianli minutiaebasedweightingaggregationofdeepconvolutionalfeaturesforveinrecognition |