Hide metadata

dc.contributor.authorLyseggen, Thomas Fossøy
dc.date.accessioned2019-08-08T23:45:45Z
dc.date.available2019-08-08T23:45:45Z
dc.date.issued2019
dc.identifier.citationLyseggen, Thomas Fossøy. Keystroke Audio as a Biometric for Authentication. Master thesis, University of Oslo, 2019
dc.identifier.urihttp://hdl.handle.net/10852/69007
dc.description.abstractPeople are bad at making passwords, resulting in an easily exploitable security vulnerability across the digital landscape. Basing your memorized password on something like your spouse’s name or even your own is inherently and intuitively foolish, but people do it regardless. The trade-off between usability and increased security is too great, and better passwords often goes on the favour of usability. Good usability is indeed important, and the ideal solution would be to maintain the usability while at the same time increasing the theoretic security. By introducing a second factor in the authentication process, the security can be significantly increased. With usability in mind, a suitable possibility is the use of keystroke dynamics as a biometric for authentication in two-factor authentication. This does not interfere with the already established authentication process through passwords. By the means of including keystroke dynamics, researchers have obtained promising results. In this study, we explore the possibility of a user’s keystroke dynamics being embedded in keystroke audio, to attempt to improve on this approach. This auditory data is acquired through the use of multiple microphones.eng
dc.language.isoeng
dc.subjectBiometrics
dc.subjectLSTM
dc.subjectbase-rate fallacy
dc.subjectprofiling
dc.subjectkeystroke dynamics
dc.subjectconvlstm
dc.subjectaudio
dc.subjectmachine learning
dc.subjectneural network
dc.subjectkeystroke
dc.subjectauthentication
dc.titleKeystroke Audio as a Biometric for Authenticationeng
dc.typeMaster thesis
dc.date.updated2019-08-08T23:45:45Z
dc.creator.authorLyseggen, Thomas Fossøy
dc.identifier.urnURN:NBN:no-72159
dc.type.documentMasteroppgave
dc.identifier.fulltextFulltext https://www.duo.uio.no/bitstream/handle/10852/69007/14/master.pdf


Files in this item

Appears in the following Collection

Hide metadata