Application of Machine Learning to Biometric Systems- A Survey

Chato, Lina; Latifi, Shahram

doi:10.1088/1742-6596/1098/1/012017

Cited by 6 publications

(5 citation statements)

References 1 publication

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Machine learning allows systems to perform tasks without being explicitly programmed to do so. Machine learning is therefore being widely used in areas including image processing and biometrics, as it can effectively analyze and interpret large datasets [39]. Machine learning models such as regression models are being used to predict the patterns in the data and generate output based on the identified patterns, or to make decisions using classifiers and pattern recognition models.…”

Section: Biometric Authenticationmentioning

confidence: 99%

“…Biometric authentication has been discussed in [38], [39] in which ECG data from hospitals and security check points were analyzed for authentication purposes. The next stage involves dividing the data into subintervals based on the peak-to-peak levels of the ECG signals with a time window determined based on the minimum heartbeat of a certain heart rate to ease the computational process.…”

Section: Biometric Authenticationmentioning

confidence: 99%

See 1 more Smart Citation

Empirical Studies of TESLA Protocol: Properties, Implementations, and Replacement of Public Cryptography Using Biometric Authentication

et al. 2022

View full text Add to dashboard Cite

This study discusses the general overview of Timed Efficient Stream Loss-tolerant Authentication (TESLA) protocol, including its properties, key setups, and improvement protocols. The discussion includes a new proposed two-level infinite µTESLA (TLI µTESLA) protocol that solves the authentication delay and synchronization issues. We theoretically compared TLI µTESLA with the previously proposed protocols in terms of security services and showed that the new protocol prevents excessive use of the buffer in the sensor node and reduces the DoS attacks on the network. In addition, it accelerates the authentication process of the broadcasted message with less delay and assures continuous receipt of packets compared to previous TESLA Protocols. We also addressed the challenges faced during the implementation of TESLA protocol and presented the recent solutions and parameter choices for improving the efficiency of the TESLA protocol. Moreover, we focused on utilizing biometric authentication as a promising approach to replace public cryptography in the authentication process.INDEX TERMS Biometric authentication, lightweight cryptography, machine learning, TESLA protocol. I. INTRODUCTION TO LIGHTWEIGHT CRYPTOGRAPHYCurrently, the internet of things (IoT) is rapidly expanding and being applied to several fields, such as in healthcare monitoring, environmental monitoring, smart censoring, and vital decision-making in different professional careers. However, the challenging features of IoT include their involvement in constrained devices such as RFIDs, sensor devices, and mobile phones, which have limited energy resources, communication bandwidth, and memory storage. With the increase in the application of these IoT devices, they become vulnerable to malicious attacks, and thus, the implementation of efficient yet lightweight security protocols is urgently needed [1].Lightweight cryptography involves simplified encryption protocols and schemes with low computational complexity that can be processed on such constrained devices to provide adequate security, considering the limited energy, bandwidth,The associate editor coordinating the review of this manuscript and approving it for publication was Mouloud Denai .

show abstract

Section: Biometric Authenticationmentioning

confidence: 99%

Section: Biometric Authenticationmentioning

confidence: 99%

Empirical Studies of TESLA Protocol: Properties, Implementations, and Replacement of Public Cryptography Using Biometric Authentication

et al. 2022

View full text Add to dashboard Cite

show abstract

“…"Who owns information?" as Ann Branscomb asked in 1994 [30], has taken on new meaning since the advent of Web 2.0, 5G mobile, advances in storage area networks and new ways to analyze the data collected through emergent approaches in artificial intelligence and machine learning (ML) [68]. Soft biometrics, which provides an additional layer of description, can also infer a great deal about someone using qualitative attributes, by capturing data about the way an individual might style their hair or wear a beard; use hats, scarfs, or eye glasses; and even capture data about their attire including the type, brand, and color of clothes; the make-up applied; other apparel adorned such as earrings and necklaces, and much more [31].…”

Section: Citizen Rights Public Spaces and Informed Consentmentioning

confidence: 99%

Biometrics and AI Bias

Michael

Jayashree

Bandara

et al. 2022

IEEE Trans. Technol. Soc.

View full text Add to dashboard Cite

“…Dasar untuk setiap metode biometrik adalah beberapa jenis metode pencocokan yang biasanya satu lawan satu dalam hal verifikasi dan satu ke banyak dalam hal identifikasi. Di sisi lain, metode Klasifikasi Deep Learning telah terbukti meningkatkan kinerja sistem biometrik dan potensi penuhnya dalam membantu biometrik mencapai kinerja 100% belum dieksplorasi secara memadai [3].…”

Section: Pendahuluanunclassified

Implementasi Jaringan Hierarki Attention Untuk Klasifikasi Basis Data Multimodal Biometrik

Purba

Sari

2022

jipi. jurnal. ilmiah. penelitian. dan. pembelajaran. informatik

View full text Add to dashboard Cite

Pada era digital ini, Teknologi semakin pesat termasuk di bidang Machine Learning, bekerja pada Machine Learning (ML) telah dipercepat dalam beberapa tahun terakhir. Klasifikasi dapat dilakukan dengan banyak metode, salah satunya yaitu menggunakan Deep Learning. Deep Learning merupakanbagian dari Machine Learning yang dapatmempelajari metode komputasinya sendiri. Deep Learning yang saat ini memilikihasil paling signifikan dalam pengenalan citra adalah Convolutional Neural Network (CNN) salah satunya sistem identifikasi menjadi hal yang penting untuk dapat mengenali seseorang berdasarkan ciri yang dimilikinya. Pada umumnya sistem identifikasi seseorang masih dilakukan menggunakan teknik konvensional, seperti menggunakan ID card yang mudah ruak dan hilang, kata sandi atau PIN yang mudah dilupakan. Saat ini, biometrik banyak digunakan untuk verifikasi dan identifikasi. Keunikan dan ketersediaan data biometrik membuatnya ideal untuk tujuan ini. Informasi biometrik biasanya mencakup fitur wajah, sidik jari, iris mata, tanda tangan, suara, dan DNA. Dalam hal ini erat kaitannya dengan sistem otentikasi biometrik multimodal yang terbukti menjadi pendekatan yang menjanjikan untuk mengidentifikasi seseorang menggunakan fitur fisiologis dan/atau perilaku seperti wajah, sidik jari, sidik jari, geometri tangan, iris, retina, vena, serta ucapan. Biometrik multimodal wajah, telapak tangan, telapak tangan dan audio memiliki beberapa keunggulan termasuk pemanfaatan akuisisi gambar non-invasif dan berbiaya rendah. Selain itu, gambar wajah dan sidik jari dapat dengan mudah diperoleh secara bersamaan menggunakan dua sensor tanpa sentuhan. Pada penelitian ini dalam klasifikasinya menggunakan jaringan Hierarki Attention sehingga memperoleh Hasil dari evaluasi matrik dari implementasi jaringan hierarki attention untuk klasifikasi basis data multimodal biometrik diperoleh akurasi sebesar 100%

show abstract

Application of Machine Learning to Biometric Systems- A Survey

Cited by 6 publications

References 1 publication

Empirical Studies of TESLA Protocol: Properties, Implementations, and Replacement of Public Cryptography Using Biometric Authentication

Empirical Studies of TESLA Protocol: Properties, Implementations, and Replacement of Public Cryptography Using Biometric Authentication

Biometrics and AI Bias

Implementasi Jaringan Hierarki Attention Untuk Klasifikasi Basis Data Multimodal Biometrik

Contact Info

Product

Resources

About