Margit Antal

Margit Antal
Sapientia Hungarian University of Transylvania · Department of Mathematics and Informatics

PhD

About

38
Publications
13,056
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
402
Citations
Additional affiliations
September 2016 - present
Sapientia Hungarian University of Transylvania
Position
  • Professor (Associate)
September 2001 - September 2016
Sapientia Hungarian University of Transylvania
Position
  • Lecturer

Publications

Publications (38)
Conference Paper
This paper investigates gender recognition from keystroke dynamics data and from touchscreen swipes. Classification measurements were performed using 10-fold cross-validation and leave-one-user-out cross-validation (LOUOCV). We show that when the target is unseen user data classification, only the second approach is viable. Based on our limited dat...
Conference Paper
The current proliferation of mobile devices raises an important issue: that of intelligent authentication mechanisms based on sensors within handheld devices. On-line signatures can now be easily captured with the fingertip on mobile devices. We study the performance of on-line signature authentication systems on touchscreen-based mobile devices. P...
Conference Paper
Full-text available
Currently people store more and more sensitive data on their mobile devices. Therefore it is highly important to strengthen the existing authentication mechanisms. The analysis of typing patterns, formally known as keystroke dynamics is useful to enhance the security of password-based authentication. Moreover, touchscreen allows adding features ran...
Article
The aim of this study is to analyze information that can be revealed from simple touch gestures such as horizontal and vertical scrolling. Touch gestures contain identity information, they can reflect the user’s experience using touchscreen and they can infer the gender of the user. The statements are based on measurements on a large touch dataset...
Chapter
In this paper we study keystroke dynamics as an authentication mechanism for touchscreen based devices. A data collection application was designed and implemented for Android devices in order to collect several types of password. Besides easy and strong passwords we propose a new type of password—logical strong—which is a strong password, but easy...
Chapter
Due to its prominent applications, time series classification is one of the most important fields of machine learning. Although there are various approaches for time series classification, dynamic time warping (DTW) is generally considered to be a well-suited distance measure for time series. Therefore, in the early 2000s, techniques based on DTW d...
Article
Full-text available
The growing interest in bot detection can be attributed to the fact that fraudulent actions performed by bots cause surprisingly high economical damage. State-of-the-art bots aim at mimicking as many as possible aspects of human behavior, ranging from response times and typing dynamics to human-like phrasing and mouse trajectories. In order to supp...
Preprint
Full-text available
Recent advances in pattern matching, such as speech or object recognition support the viability of feature learning with deep learning solutions for gait recognition. Past papers have evaluated deep neural networks trained in a supervised manner for this task. In this work, we investigated both supervised and unsupervised approaches. Feature extrac...
Article
Full-text available
Behavioural biometrics provides an extra layer of security for user authentication mechanisms. Among behavioural biometrics, mouse dynamics provides a non-intrusive layer of security. In this paper we propose a novel convolutional neural network for extracting the features from the time series of users’ mouse movements. The effect of two preprocess...
Article
Full-text available
Compared to other behavioural biometrics, mouse dynamics is a less explored area. General purpose data sets containing unrestricted mouse usage data are usually not available. The Balabit data set was released in 2016 for a data science competition which, despite the few users covered, can be considered the first adequate publicly available data se...
Conference Paper
Full-text available
Portable sensor-based gait biometrics has been actively researched in recent years. Generative models are rarely examined as part of gait recognition systems and there are several open questions with regards to machine learning approaches in this context. This paper shows that cross-day evaluations and the selection method of impostors influence sy...
Preprint
Full-text available
Compared to other behavioural biometrics, mouse dynamics is a less explored area. General purpose data sets containing unrestricted mouse usage data are usually not available. The Balabit data set was released in 2016 for a data science competition, which against the few subjects, can be considered the first adequate publicly available one. This pa...
Article
Full-text available
We present MOBISIG, a pseudosignature dataset containing finger-drawn signatures from 83 users captured with a capacitive touchscreen-based mobile device. The database was captured in three sessions resulting in 45 genuine signatures and 20 skilled forgeries for each user. The database was evaluated by two state-of-the-art methods: a function-based...
Article
Full-text available
The widespread use of smartphones and the ability of these devices to digitize signatures have made it possible to sign electronic documents in this way. In this paper we compared two on-line signature databases in terms of verification performance: the MCYT containing signatures drawn by stylus pen, and MOBISIG containing finger drawn signatures....
Conference Paper
Recently a new set consisting of six information theory features was proposed for on-line signature verification by Rosso, Ospina and Frery. The proposed features were evaluated on the MCYT-100 on-line signature database resulting in the best performance ever measured on that dataset. In this paper we repeat their measurements and show that their r...
Data
Full-text available
Each user has a dedicated folder which contains the 45 genuine signatures of the user and the 20 forgeries made by other users. The naming convention of the files is as follows: SIGN_[TYPE]_USER[SID]_USER[WID]_[NR].csv where  TYPE is FOR for forgeries and GEN for genuine signatures.  SID is the identifier of the user whose signatures contains the...
Conference Paper
Several types of information can be revealed from data provided by mobile sensors. In this study touchscreen and accelerometer data was collected from a group of 98 volunteers during filling in the Eysenck Personality Questionnaire on a tablet computer. Subjects performed swipes on the touchscreen in order to answer the questions. Touchscreen swipe...
Conference Paper
Full-text available
In this work we investigated user authentication on mobile devices using touch behavior and micro movements of the device. The novelty of our work lies in the collection of user behavior data during the filling in of a psychological questionnaire (implemented as an Android application). In order to answer the questions, users were required to use a...
Conference Paper
In this paper we study keystroke dynamics as an authentication mechanism for touchscreen based devices. The authentication process decides whether the identity of a given person is accepted or rejected. This can be easily implemented by using a two-class classifier which operates with the help of positive samples (belonging to the authentic person)...
Article
Full-text available
Currently people store more and more sensitive data on their mobile devices. Therefore it is highly important to strengthen the existing authentication mechanisms. The analysis of typing patterns, formally known as keystroke dynamics is useful to enhance the security of password-based authentication. Moreover, touchscreen allows adding features ran...
Data
Data collection Number of subjects: 42 (24 male, 18 female) Password: .tie5Roanl Number of samples/subject: 51 Number of sessions: at least 2 sessions/user Devices: 37 subjects - Nexus 7 tablet, 5 subjects - Mobil LG Optimus L7 II P710 (user ids: 6, 7, 8, 9, 10)
Article
Full-text available
Due to the powerful sensors incorporated, the new generations of smartphones have become capable of many sophisticated biometrics. Touchscreen based biometrics is a new type of biometrics showing great potential. In this paper we review the studies already conducted in this direction, then present our study aimed to find the best method for touch d...
Data
Data collection 71 subjects (56 male, 15 female) 8 different Android devices (resolutions: 320x480 ..1080x1205)) personal information: gender, birth date, touch screen experience level (0,1,2,3) vertical strokes: text comprehension horizontal strokes: image gallery - choosing the favourite image multiple sessions Feature extraction strokeDuratio...
Article
Full-text available
In this paper, we present a new item response model for computerized adaptive testing: Item Response Theory combined with Elo rating. Adaptive test systems require a calibrated item bank and item calibration methods are usually based on Item Response Theory (IRT). However, these methods require item pretesting on large sample sizes, which is very e...
Article
Full-text available
One of the fastest evolving field among teaching and learning research is students' performance evaluation. Computer based testing systems are increasingly adopted by universities. However, the implementation and maintenance of such a system and its underlying item bank is a challenge for an inexperienced tutor. Therefore, this paper discusses the...
Conference Paper
Full-text available
The aim of this study is to answer two questions regarding the use of phonetic information for speaker modelling. We formulate answers for (1) what are the dis-criminative powers of broad phonetic classes for the task of speaker identification? (2) Are the phonetic speaker models more suitable for speaker recognition than standard models?
Article
Full-text available
This paper presents a simple speech recognition system using Gaussian mixtures as phoneme models. The proposed architecture does not follow the integrated search strategy. Instead we use a modular design. We propose two modiflcations to the Viterbi decoding algorithm in order to be applicable to our phoneme models. Both strategies have been imple-...
Conference Paper
The aim of this study is to provide a quantitative assessment of the speaker discriminating properties of broad phonetic groups. GMM based approach to speaker modelling is used in conjunction with a phonetically hand-labelled speech database (TIMIT) to produce broad phonetic group ranking based on speaker identification scores. The broad phonetic g...
Article
Vector quantisation and Gaussian mixture modelling methods are very popular methods for automatic speaker identification. First we give a concise overview of these methods, then present some measurements com-paring them on behalf of the TIMIT corpus. The aim of this paper is to study the influence of the speech material on performances of such meth...
Article
Full-text available
This paper examines statistical models for phoneme classica- tion. We compare the performance of our phoneme classication system using Gaussian mixture (GMM) phoneme models with systems using hidden Markov phoneme models (HMM). Measurements show that our model's per- formance is comparable with HMM models in context independent phoneme classication...
Article
We compare recognition performance of Vector Quantization method (VQ) and Gaussian Mixture Modeling method (GMM) in normal speech conditions. We perform measurements to emphasize the relationship between the size of models – number of clusters or number of components of mixtures and size of the system. We also conclude that the VQ method is a parti...
Article
In this paper the GMM speaker model was analyzed from the viewpoint of its phonetic content. Phoneme distribution among clusters represented by Gaussians was studied. Special speaker models were also created using only a part of the training data, in order to identify the most valuable part of speech, for the purpose of speaker identification.

Questions

Questions (2)
Question
This data set is described in an article published in Computers & Security, 45, pp. 156-171, 2014.
"A public benchmark mouse-dynamics data set is established, not only for this study, but also to foster future research."
The following link published in the article does not work:
I have already sent a few e-mails to Chao Shen, but I got no answer.
Question
Besides classification accuracy I want to plot ROC and DET curves.  

Network

Cited By

Projects

Projects (2)
Project
The objective of this project is the development of unobtrusive biometric systems using exclusively the embedded sensors that can protect sensitive data stored on mobile devices.
Project
http://www.ms.sapientia.ro/~manyi/mobisig.html