Finger-Vein Pattern Restoration With Generative Adversarial Network

Yang, Shuqiang; Qin, Huafeng; Liu, Xia; Wang, Jun

doi:10.1109/access.2020.3009220

Cited by 8 publications

(2 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The various CNN architectures were introduced in [15] for finger vein matching performance. A novel finger-vein image restoration approach was introduced based on a generative adversarial network (GAN) [16]. Though the move was intended to improve verification accuracy, the time required for image restoration was not reduced.…”

Section: Related Workmentioning

confidence: 99%

Deep Belief Structure Learning for Finger Vein Verification by Solving the Biometric Recognition Model

Muthusamy¹,

Ponnusamy²

2022

Preprint

View full text Add to dashboard Cite

Vein authentication is a novel biometric method to authenticate the individuality of a person. The conventional biometric technique employs shape images and exact segments of finger veins for the verification process. To improve the verification accuracy, a novel Anisotropic Filtered Stromberg Feature Transform based on Tucker’s Congruence Deep Belief Structure Learning (AFSFT-TCDBSL) technique is intended. The proposed AFSFT-TCDBSL technique comprises one input, three hidden, and one output layers. The numbers of images are collected in the input, and input images are preprocessed using anisotropic diffusion filtering in the first hidden layer. Finally, the verification process is performed using Tucker’s congruence correlation coefficient (TCCC). Based on the correlation, the verification outputs are getting to the output layer. In this way, accurate finger vein verification is performed with superior accuracy and with a minimum false rate. We performed experimental assessments with different factors, such as PSNR, FVVA, FPR, and CT. The proposed ADFSFT-TCDBSL technique offers better finger-vein verification results than the state-of-the-art methods.

show abstract

Section: Related Workmentioning

confidence: 99%

Deep Belief Structure Learning for Finger Vein Verification by Solving the Biometric Recognition Model

Muthusamy¹,

Ponnusamy²

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Currently, with the development of deep learning, Convolutional Neural Networks (CNN) [10], Deep Belief Networks (DBNS) [11], and Generative Adversarial Networks(GAN), have been used to learn robust features from raw pixel images, and have shown good performance. Das et al [12] developed a finger vein recognition system based on a CNN, which is capable of processing finger vein images of varying quality while obtaining stable and highly accurate recognition performance. Boucherit et al [13] proposed a merged convolutional neural network, which merged multiple short-path CNN structures, to extract the features from images of varying quality and fuse them.…”

Section: Introductionmentioning

confidence: 99%

FV-DMHN: Dual Multi-Head Network for Finger Vein Recognition

An,

Ren,

Tao

2024

IEEE Access

View full text Add to dashboard Cite

Deep learning-based finger vein image recognition methods usually suffer from high complexity, insufficient global information extraction, and overfitting. The use of lightweight networks can significantly reduce the accuracy owing to the reduction in model parameters. For this reason, this paper proposes a Dual Multi-Head neural Network for Finger Vein Recognition (FV-DMHN), which combines the Multi-Head Self-Attention (MHSA) mechanism with the Multi-Head Convolutional Network(MHCN)to increase the training efficiency of the network while expanding the CNN horizon. The inverted residual structure is also used in the network to enhance the expressive power of the network. The algorithm achieves recognition accuracies of 99.81%, 99.67%, 99.69%, and 99.83% on three publicly available datasets, FV-USM, SDUMLA-HMT, THU-FVFDT2, and self-built datasets FV-SIPL, respectively, with an average equal-error rate of 0.339%, and the recognition time of a single image is only 2.63 ms. The experimental results show that the algorithm is superior to other methods in terms of accuracy and average equal error rate, at the same time, it not only reduces the number of network parameters and computational complexity but also achieves excellent recognition speed.

show abstract