| | SLO | ENG | Cookies and privacy

Bigger font | Smaller font

Show document Help

Title:Dejavniki kredibilnosti sintetične vsebine : diplomsko delo visokošolskega študijskega programa Informacijska varnost
Authors:ID Kalšan, Domagoj (Author)
ID Vrhovec, Simon (Mentor) More about this mentor... New window
Files:.pdf VS_Kalsan_Domagoj_2022.pdf (1,57 MB)
MD5: CF796E902ECFF620647FF14065EFE84C
PID: 20.500.12556/dkum/8104b477-6b69-44a1-9b56-a3c26d66de36
 
Language:Slovenian
Work type:Bachelor thesis/paper
Typology:2.11 - Undergraduate Thesis
Organization:FVV - Faculty of Criminal Justice and Security
Abstract:Sintetična vsebina krepi svojo prisotnost v digitalnem prostoru že nekaj let. Vsakodnevno raziskovalci z celega sveta odkrijejo nove načine ustvarjanja dovršene sintetične vsebine. Na različne načine želijo ustvariti popolno sintetično vsebino, ki jo uporabniki ne zmorejo ločiti od avtentične vsebine. Sintetična vsebina v svoji osnovi ni škodljiva in se odlično uporablja za popravljanje končnih izdelkov. Takšne tehnologije se že dolgo poslužuje filmska industrija, oglaševalske in tržne agencije ter drugi. Uporaba algoritmov za ustvarjanje sintetične vsebine zmanjša ceno in čas, ki sta potrebna za ustvarjanje dovršenih posnetkov. Zaradi učinkovitosti sintetične vsebine in njene lahke uporabe se je sintetična vsebina hitro začela uporabljati tudi v škodljive namene. V veliki meri je prizadela znane osebnosti in predvsem ženske, saj se je njena uporaba hitro razširila v industrijo za odrasle. Poleg uničevanja ugleda posameznikov je sintetična vsebina velikokrat uporabljena v politične namene in ustvarjanja kaosa. Zaradi povečanega števila škodljive sintetične vsebine so začele korporacije in ustanove pospešeno razvijati algoritme, ki bi bili sposobni prepoznati sintetično vsebino. Ker tudi uporabniki, ki so seznanjeni z obstojem sintetične vsebine ugibajo pri njeni prepoznavi, smo želeli ugotoviti človeško sposobnost prepoznave sintetične vsebine. Področje človeške prepoznave sintetične vsebine je premalo raziskano. Da bi ugotovili kateri dejavniki vplivajo na človeško prepoznavo sintetične vsebine smo izvedli kvazi eksperiment. Anketiranci so bili na kratko seznanjeni z sintetično vsebino in so poizkušali pravilno ugotoviti v katerih posnetkih je bila prisotna sintetična vsebina. Naše delo je le majhen korak v ugotavljanje dejavnikov kredibilnosti sintetične vsebine, z njem pa želimo povečati zanimanje za to področje.
Keywords:diplomske naloge, sintetična vsebina, manipulacija posnetkov, nevronske mreže
Place of publishing:Ljubljana
Place of performance:Ljubljana
Publisher:[D. Kalšan]
Year of publishing:2022
Year of performance:2022
Number of pages:VII, 35 str.
PID:20.500.12556/DKUM-82103 New window
UDC:343.72:004(043.2)
COBISS.SI-ID:117225987 New window
Publication date in DKUM:03.08.2022
Views:864
Downloads:126
Metadata:XML DC-XML DC-RDF
Categories:FVV
:
Copy citation
  
Average score:(0 votes)
Your score:Voting is allowed only for logged in users.
Share:Bookmark and Share



Hover the mouse pointer over a document title to show the abstract or click on the title to get all document metadata.

Licences

License:CC BY-NC-ND 4.0, Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International
Link:http://creativecommons.org/licenses/by-nc-nd/4.0/
Description:The most restrictive Creative Commons license. This only allows people to download and share the work for no commercial gain and for no other purposes.
Licensing start date:13.07.2022

Secondary language

Language:English
Title:Factors of deepfake credibility
Abstract:Deepfakes are becoming very common in social media. Researchers from around the globe are figuring out new ways of creating perfect deepfakes. They are trying to create detailed deepfakes which users are not able to distinguish from authentic video. Deepfakes in general are very useful and have been used for a long time to complete video scenes. Film industry and marketing corporations have been using deepfakes to their advantage for quite some time now. Using deepfakes for film production greatly decreases time and cost to produce minor adjustments. Because it has become increasingly easy to create deepfakes people have started to use them for malicious purposes. Deepfakes have quickly become known for their use in defamation of celebrities, creation of adult video content and creating user distrust. The increase in deepfake creation has also increased awareness of malicious implications of deepfakes and encouraged corporations to begin creating deepfake detection algorithms. The problem with deepfakes is that even users who are acquainted with such technology are unable to discern between authentic video and deepfakes. The field of user detection of deepfakes is poorly researched and there is only a handful of research done on factors of deepfake credibility. To better understand human perception of deepfakes we have conducted a quasi-experiment. First, we introduced our respondents to deepfakes and then they were shown a set of videos. We asked them to discern if the video they saw is fake or real. Our work is a small step towards understanding of human perception on deepfakes. With this research we are trying to figure out factors of deepfake credibility and encourage further research on this topic.
Keywords:Deepfakes, video manipulation, neural networks


Comments

Leave comment

You must log in to leave a comment.

Comments (0)
0 - 0 / 0
 
There are no comments!

Back
Logos of partners University of Maribor University of Ljubljana University of Primorska University of Nova Gorica