Fingerprint authentication is a widely used method to verify someone's identity by analysing unique fingerprint features, such as ridges and specific points called minutiae. However, there are concerns about its vulnerability to fake fingerprints and privacy issues. Cancellable biometrics is a promising solution to tackle these concerns. It transforms fingerprint features into secure forms that cannot be reversed back to the original, even if someone gets hold of them. This paper proposes an alignment-free secure fingerprint authentication method that integrates minutiae point descriptors and Scale Invariant Feature Transform (SIFT) keypoint descriptors, enhanced with Elliptic Curve signcryption, aiming to fortify security without compromising authentication accuracy. Experimental evaluations were conducted using the Fingerprint Verification Competition (FVC) 2002 dataset, showcasing the efficacy of the proposed approach. Experimental results demonstrate a significant reduction in security risks while upholding authentication accuracy, thus affirming the effectiveness of our methodology in enhancing fingerprint authentication security.

Multimodal biometric systems have gained popularity for their enhanced recognition accuracy and resistance to attacks like spoofing. In this paper, we introduce a novel approach to safeguard multimodal biometric templates using a Cancelable Convolutional Neural Network (CCNN). Our method utilizes two biometric traits, the iris and fingerprint. Initially, features are extracted separately from these traits and then combined into a single feature vector. Subsequently, a CCNN is applied to reduce the size of this fused vector. Finally, the reduced vector is multiplied with a user-provided seed for enhanced cancelability. Evaluations on the Children Multimodal Biometric Database (CMBD), CASIA Iris V3, and FVC 2002 DB2 demonstrate that our method effectively balances user privacy and accuracy while maintaining a high level of precision. With an exceptionally low Equal Error Rate (EER) of 0.073% and 0.038% on both datasets. Our method fulfills the requirements of diversity, irreversibility, and revocability showcasing its efficiency in terms of security and accuracy.

The use of a biometric authentication system (BAS) for reliable automatic human recognition has increased exponentially over traditional authentication systems in recent years. Multimodal BAS was introduced to solve unimodal BAS's difficulties and improve security. Privacy and security are two significant concerns to be addressed in BAS, as biometric traits are irrevocable. Researchers employed cancelable biometrics in the past few years to propose several privacy-preserving BAS. We propose a privacy-preserving bimodal cancelable BAS (PPBCBAS) to overcome these problems. The traits used in our method are iris and fingerprint. Features are extracted from both the traits, and feature level fusion is done by concatenating the feature vectors of iris and fingerprint. PPBCBAS uses a quotient filter to generate the cancelable template, and the comparison is made on these transformed templates using the modified Hamming distance. PPBCBAS has been tested on three publicly available databases to analyze its efficiency. PPBCBAS satisfies the diversity, irreversibility, and revocability properties and achieves decent performance.

In this paper multimodal biometric system is developed using two traits iris and fingerprint. The features generated by iris and fingerprint images are fused at the feature level. The generated fused feature vector template cannot be stored directly on the server, if stored directly can lead to various privacy and security concerns. So, these templates are encrypted in such a way that even when applying any operations on the templates, the templates should be in encrypted form. So, the operations need to be performed in the encrypted domain without decrypting it, and the final result, when decrypted should again give back the correct result as if the operations are performed on the original data. Fully Homomorphic encryption (FHE) scheme is designed to satisfy the above conditions. FHE is used to compute the hamming distance between the reference and probe template in an encrypted domain. To improve accuracy rotational invariant technique is used, which solves rotational inconsistency problems. The computational speed is increased by using a batching scheme to reduce the number of operations during homomorphic multiplication. We have conducted our experiment on the IITD and CASIA dataset. The best EER is obtained in CASIA dataset of 0.01% with a computational time of 0.0152 sec per template.

In the ongoing years, biometric systems end up helpless against the spillage of template information. If a biometric template is stolen, it is lost permanently and cannot be restored or reissued. Here, we use iris biometric because of its high accuracy. In this paper, we develop a new cancellable biometric scheme using the indexing-first-one (IFO) hashing coupled with a technique called partial sort. The IFO hashing uses new mechanisms called the P-order Hadamard product and modulo threshold function paired with the partial sort technique which has considerably strengthened it further. We used the very sophisticated CASIA-v3 database which provides us with a wide range of iris templates for our experiments. As compared to the previous cancellable schemes, the analysis of the results of these experiments provides us with good accuracy and strong resistance to various privacy and security attacks.

Authentication systems are now an important part of our daily life. Human biological, behavioral, and physical characteristics are usually applied in authenticating a person in various applications. Unimodal biometric systems have a number of limitations, such as noise sensitivity, population coverage, intra-class variations, non-universality, and vulnerability to spoofing. Multimodal biometric systems overcome these limitations and are being widely used in many real-world applications. In this work, to construct a three-dimensional (3-D) shell, we use fingerprint and iris. First, features are extracted from the fingerprint. Then, using a user key set, a two-dimensional spiral curve is generated from fingerprint features. Next, iris features are extracted using a pre-trained VGG-16 model, then feature vector-based random projection is applied to generate an iris feature vector. This generated feature vector is combined with the fingerprint shell to construct a secured 3-D shell. Finally, these fused 3-D templates are saved in the database and are used for matching. Our proposed technique has been evaluated on the three publicly available datasets, showing that it can preserve user privacy while maintaining the accuracy of the system with an equal error rate of 0.09%, 0.032%, and 0.015%.

Deep learning shown its potential in a variety of medical applications and proved as a count on by people as a step ahead approach compared to traditional machine learning models. Moreover, the other implementations of these models such as the convolutional neural networks (CNNs) provide extensive applications in the field of medicine, which usually involves processing and analysis of a large dataset. This paper aims to create a CNN model which can solve the problem of white blood cell subtyping which is a daunting one in clinical processing of blood. The manual classification of white blood cells in laboratory is a time-consuming process which gives rise to the need for an automated process to perform the task. A CNN-based machine learning model is developed to classify the leukocytes into their proper subtypes by performing tests on a dataset of around twelve thousand images of leukocytes and their types, and a wide range of parameters is evaluated. This model can automatically classify the white blood cells to save manual labor, time and improve efficiency. Further, pretrained models like Inception-v3, VGGNet and AlexNet are used for the classification, and their performance is compared and analyzed.

Due to the rapid advancement of technology, biometrics is widely used in authenticating a person. The primary issue in biometric authentication is template protection, which protects the user's privacy. To achieve this level of security, we transform the original template generated from the extracted features into a pseudo template. This paper considers two biometric traits, namely Iris and Fingerprint. First, the features are extracted from both the traits. After extracting the features, the feature vectors are compressed to reduce the size. Then the feature vectors of both the traits are fused. Later, local random projection (LRP) is applied to this template to build a revocable and un-linkable template. Finally, fully homomorphic encryption (FHE) is applied to this generated template to secure the user's privacy, as all operations on the template are carried on encrypted data. Our proposed method improves the accuracy of the system as it generates rotational invariant codes. This proposed method is tested on the Children Multimodal Biometric Database (CMBD), which shows that our method can maintain user privacy with a good accuracy rate. Our method achieves an EER of 0.0214%.

Iris recognition technologies applied to produce comprehensive and correct biometric identification of people in numerous large-scale data of humans. Additionally, the iris is stable over time, i.e., iris biometric knowledge offers links between biometric characteristics and people. The e-business and e-governance require more machine-driven iris recognition. It has millions of iris images that are in near-infrared illumination. It is used for people’s identity. A variety of applications for surveillance and e-business will embody iris pictures that are unit non-heritable below visible illumination. The self-learned iris features are created by the convolution neural network (CNN), give more accuracy than handcrafted feature iris recognition. In this paper, a modified iris recognition system is introduced using deep learning techniques along with multi-class SVM for matching. We use the Poly-U database, which is from 209 subjects. CNN with softmax cross-entropy loss gives the most accurate matching of testing images. This method gives better results in terms of EER. We analyzed the proposed architecture on other publicly available databases through various experiments.