One-step, three-factor authentication in a single earpiece

Curran, Max T.; Merrill, Nick; Chuang, John; Gandhi, Swapan

doi:10.1145/3123024.3123087

Cited by 10 publications

(11 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…(3) The paper had to propose an authentication scheme. Specifically, the paper had to use machine learning to label [12], [33], [41], [42] IMWUT/UbiComp [11], [20], [25], [35], [48] INFOCOM [8], [36], [45], [51], [53] MobiCom [15], [32] MobiSys [7], [31] NDSS [5], [17], [49], [50] Pattern Recognition [2], [9], [18], [24], [38], [40], [54]…”

Section: Review Of Recent Authentication Systemsmentioning

confidence: 99%

Robust Performance Metrics for Authentication Systems

Sugrim

Liu

McLean³

et al. 2019

Proceedings 2019 Network and Distributed System Security Symposium

View full text Add to dashboard Cite

Research has produced many types of authentication systems that use machine learning. However, there is no consistent approach for reporting performance metrics and the reported metrics are inadequate. In this work, we show that several of the common metrics used for reporting performance, such as maximum accuracy (ACC), equal error rate (EER) and area under the ROC curve (AUROC), are inherently flawed. These common metrics hide the details of the inherent tradeoffs a system must make when implemented. Our findings show that current metrics give no insight into how system performance degrades outside the ideal conditions in which they were designed. We argue that adequate performance reporting must be provided to enable meaningful evaluation and that current, commonly used approaches fail in this regard. We present the unnormalized frequency count of scores (FCS) to demonstrate the mathematical underpinnings that lead to these failures and show how they can be avoided. The FCS can be used to augment the performance reporting to enable comparison across systems in a visual way. When reported with the Receiver Operating Characteristics curve (ROC), these two metrics provide a solution to the limitations of currently reported metrics. Finally, we show how to use the FCS and ROC metrics to evaluate and compare different authentication systems.

show abstract

Section: Review Of Recent Authentication Systemsmentioning

confidence: 99%

Robust Performance Metrics for Authentication Systems

Sugrim

Liu

McLean³

et al. 2019

Proceedings 2019 Network and Distributed System Security Symposium

View full text Add to dashboard Cite

show abstract

“…• Once an accepting sample via the feature vector API has been found, it may be possible to obtain an input that results in this sample (after feature extraction), as demonstrated by Garcia et al with the training of an autoencoder for both feature extraction and the regeneration of the input image [23]. • In this work, we have focused on authentication as a binary classification problem, largely because of its widespread use in biometric authentication [8], [9], [10], [11], [12], [13], [14], [15], [16], [26]. However, authentication has also been framed as a one-class classification problem [56], [26] or as multi-class classification [26], e.g., in a discrimination model, as noted earlier.…”

Section: Discussionmentioning

confidence: 99%

“…The target user's data for training is obtained during the registration or enrollment phase. For the negative class, the usual process is to use the data of a subset of other users enrolled in the system [27], [8], [7], [9], [10], [11], [12], [13], [14], [15], [16]. Following best machine learning practice, the data (from both classes) is split into a training and test set.…”

Section: A Biometric Authentication Systemsmentioning

confidence: 99%

“…These systems are mostly based on machine learning: a binary classifier is trained on the target user's data (positive class) and a subset of data from other users (negative class). This process is used to validate the performance of the machine learning classifier and hence the biometric system [8], [7], [9], [10], [11], [12], [13], [14], [15], [16]. The resulting proportion of negative samples (other users' data) successfully gaining access (when they should have been rejected) produces the false positive rate (FPR, also referred as False Acceptance Rate).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

On the Resilience of Biometric Authentication Systems against Random Inputs

Zhao¹,

Asghar²,

Kâafar³

2020

Proceedings 2020 Network and Distributed System Security Symposium

View full text Add to dashboard Cite

We assess the security of machine learning based biometric authentication systems against an attacker who submits uniform random inputs, either as feature vectors or raw inputs, in order to find an accepting sample of a target user. The average false positive rate (FPR) of the system, i.e., the rate at which an impostor is incorrectly accepted as the legitimate user, may be interpreted as a measure of the success probability of such an attack. However, we show that the success rate is often higher than the FPR. In particular, for one reconstructed biometric system with an average FPR of 0.03, the success rate was as high as 0.78. This has implications for the security of the system, as an attacker with only the knowledge of the length of the feature space can impersonate the user with less than 2 attempts on average. We provide detailed analysis of why the attack is successful, and validate our results using four different biometric modalities and four different machine learning classifiers. Finally, we propose mitigation techniques that render such attacks ineffective, with little to no effect on the accuracy of the system.

show abstract

“…An alternative is required. These two issues need to be resolved in employing three-factor authentication: to address the first issue, we need to develop a simple three-factor authentication process such as one-step three-factor authentication (Curran et al, 2017 ) and to address the second issue, biometrics must necessarily become more ubiquitous. If biometrics are popularly used and the biometric information for each user is saved in their smart device and not in the Cloud, the additional risk of biometric information leakage would not emerge.…”

Section: Workable Solutions and Future Challengesmentioning

confidence: 99%

Authentication of Patients and Participants in Health Information Exchange and Consent for Medical Research: A Key Step for Privacy Protection, Respect for Autonomy, and Trustworthiness

2018

View full text Add to dashboard Cite

Genome and other data are already being used in areas including cancer and rare diseases. Data-sharing and secondary uses are likely to become much broader and far more extensive; thus, obtaining proper consent for these new uses of data is an important issue. Obtaining consent through online methods may be an option to overcome the problems associated with one-off, paper-based informed consent. When the process of obtaining consent takes place remotely, authentication must be assured. Patients may also choose to store some of their own information online, such as genetic information, and allow healthcare professionals to access this data. In this health information transfer and exchange process, it is vital that anyone accessing this information be correctly authenticated to protect patients' privacy. In this article, we first clarified that authentication has two roles: i.e., not only to prevent impersonation but also to prove intent, which is a vital step to ensure that medical research and health information exchange are conducted ethically. We then set out methods of authentication. As a result, we were able to make suggestions about the requirements for authentication and a possible method of authentication for these purposes. We considered problems of biometrics and recommended two-factor authentication without biometrics as a workable solution. However, three-factor authentication including biometrics seems likely to be used once biometrics become more common.

show abstract

One-step, three-factor authentication in a single earpiece

Cited by 10 publications

References 9 publications

Robust Performance Metrics for Authentication Systems

Robust Performance Metrics for Authentication Systems

On the Resilience of Biometric Authentication Systems against Random Inputs

Authentication of Patients and Participants in Health Information Exchange and Consent for Medical Research: A Key Step for Privacy Protection, Respect for Autonomy, and Trustworthiness

Contact Info

Product

Resources

About