Frequency and Time Localization in Biometrics: STFT vs. CWT

Biometrics is a science discipline dealing with unique traits of the individuals including habitual characteristics. Once the unique features revealed from these characteristics are extracted as signals instead of raw data, it is possible to search for the new distinguishable traits. One of the traits, if any kind of time based signal is extracted, could be frequency component versus time component of the signal, depending on the content of the feature extraction and yet most of the signals consisting of a magnitude or a value as a dependent variable fit the requirements. The magnitude/value representation may vary indeed; however our previous researches proved that key-codes and key-press signals in biometric keystroke authentication and dislocation and speed signals in online signature verification are very useful of this kind of analysis. Therefore in this paper, we present the methods for extracting distinguishable features and frequency vs. time representation by short time Fourier transform (STFT) and continuous wavelet transformation (CWT) to introduce related research methodologies by comparing the outcomes and future opportunities. While presenting our approach to Biometric keystroke authentication and online signature verification, we also aim to present the differences of these methods with basic roadmaps and graphical results.

[1]  Issa Traoré,et al.  Improving Performance and Usability in Mobile Keystroke Dynamic Biometric Authentication , 2016, 2016 IEEE Security and Privacy Workshops (SPW).

[2]  Ondrej Krejcar,et al.  Hidden Frequency Feature in Electronic Signatures , 2016, IEA/AIE.

[3]  Orcan Alpar,et al.  Frequency spectrograms for biometric keystroke authentication using neural network based classifier , 2017, Knowl. Based Syst..

[4]  Ondrej Krejcar,et al.  Biometric Keystroke Signal Preprocessing Part I: Signalization, Digitization and Alteration , 2017, IEA/AIE.

[5]  Orcan Alpar.  Keystroke recognition in user authentication using ANN based RGB histogram technique , 2014, Eng. Appl. Artif. Intell..

[6]  Santanu Chaudhury,et al.  A Robust Online Signature Based Cryptosystem , 2014, 2014 14th International Conference on Frontiers in Handwriting Recognition.

[7]  Neyire Deniz Sarier.  Multimodal biometric Identity Based Encryption , 2018, Future Gener. Comput. Syst..

[8]  Katherine M. Malan,et al.  User Authentication based on Continuous Touch Biometrics , 2016, South Afr. Comput. J..

[9]  Orcan Alpar.  Intelligent biometric pattern password authentication systems for touchscreens , 2015, Expert Syst. Appl..

[10]  Soumik Mondal,et al.  A study on continuous authentication using a combination of keystroke and mouse biometrics , 2017, Neurocomputing.

[11]  Ondrej Krejcar,et al.  Biometric Swiping on Touchscreens , 2015, CISIM.

[12]  Arun Ross,et al.  Biometric authentication via keystroke sound , 2013, 2013 International Conference on Biometrics (ICB).

[13]  Ondrej Krejcar,et al.  Pattern Password Authentication Based on Touching Location , 2015, IDEAL.

[14]  Ondrej Krejcar,et al.  Biometric Keystroke Signal Preprocessing Part II: Manipulation , 2017, IEA/AIE.

[15]  Kevin Warwick,et al.  Non-conventional keystroke dynamics for user authentication , 2017, Pattern Recognit. Lett..

[16]  Ondrej Krejcar,et al.  Online signature verification by spectrogram analysis , 2017, Applied Intelligence.

[17]  Jacob O. Wobbrock,et al.  WalkType: using accelerometer data to accomodate situational impairments in mobile touch screen text entry , 2012, CHI.