東華大學圖書館 |

Digital Cues in Multimedia Forensics = = Digitale Spuren in der Multimedia-Forensik.

Record Type:	Electronic resources : Monograph/item
Title/Author:	Digital Cues in Multimedia Forensics =/
Reminder of title:	Digitale Spuren in der Multimedia-Forensik.
Author:	Mullan, Patrick.
Published:	Ann Arbor : ProQuest Dissertations & Theses, : 2021,
Description:	137 p.
Notes:	Source: Dissertations Abstracts International, Volume: 83-02, Section: B.
Contained By:	Dissertations Abstracts International83-02B.
Subject:	Software. -
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=28467670
ISBN:	9798744476199

Digital Cues in Multimedia Forensics = = Digitale Spuren in der Multimedia-Forensik.
Mullan, Patrick.

Digital Cues in Multimedia Forensics =Digitale Spuren in der Multimedia-Forensik. - Ann Arbor : ProQuest Dissertations & Theses, 2021 - 137 p.

Source: Dissertations Abstracts International, Volume: 83-02, Section: B.

Thesis (Ph.D.)--Friedrich-Alexander-Universitaet Erlangen-Nuernberg (Germany), 2021.

This item must not be sold to any third party vendors.

This work presents several contributions to the scientific field of digital multimedia forensics. The field addresses questions of authentication examination and source identification of multimedia files. All investigated multimedia entities are visual information carriers, i.e., images and videos. The subfield of authentication examination proposes methods which allow to verify or falsify the authenticity of the probed media object. Questions concerning source identification help to identify the source of the media file, where the term source offers different levels of granularity, for instance, the make or the model of a camera. Finding answers to questions of this kind, requires to operate with characteristic cues which suit the investigated context. Such cues, or features, can be derived from the byte stream of the file as well as from the visual content, i.e., the decoded pixels. The selected feature domain depends on the exact problem statement. Analysis on pixel data is especially revealing if the manipulation is of good quality. In contrast, applications that work with encoded files can be efficiently deployed to work on large scales of media file. Oftentimes, file-based methods are also helpful if the quality of the visual content makes working with pixel-based applications rather difficult.Therefore, this dissertation subsumes multiple of our published works from the field of digital multimedia forensics, which propose various new developed approaches for both authentication examination and source identification. Additionally, the suggested approaches work in different domains. Through this diversity, we are capable to cover a broad field of research questions and can solve different types of challenges.The first part of this work focuses on source identification. We first illustrate that the current literature does not cover developments introduced by the omnipresent usage of smartphones as imaging devices. Smartphones work with exchangeable software imaging routines. Hence, we suggest to introduce a software components axis orthogonal to a hardware components axis, i.e., proposing a new level of source identification. With the help of a large scale study on Apple devices, we demonstrate that different software versions have characteristic properties. We use this finding to predict software versions by resorting to machine learning. The influence of software when taking the image and when postprocessing is also investigated. Further, we illustrate how images from unknown models can be automatically associated with the make of the source device.The second part of the dissertation focuses on videos and their representation in modern video codecs. We discuss neural networks in terms of their ability to identify double compression in videos. Double compression is a necessary step that happens automatically when manipulating a video. The introduced method is likewise capable to predict a specific parameter, the so-called ``group of pictures'' length in the first compression of a double compressed video. We further introduce methods which are capable to locate manipulations in pixel-domain, for example, persons spliced into the footage. We demonstrate that the same method can be successfully used to probe if video sequences originate from the same video. The methods are benchmarked against state-of-the art solutions and consistently outperform them.

ISBN: 9798744476199Subjects--Topical Terms:

619355
Software.

Digital Cues in Multimedia Forensics = = Digitale Spuren in der Multimedia-Forensik.
LDR:08448nmm a2200385 4500 001 2345926
005 20220613064816.5
008 241004s2021 ||||||||||||||||| ||eng d
020 $a 9798744476199
035 $a (MiAaPQ)AAI28467670
035 $a (MiAaPQ)FAUErlangen15953
035 $a AAI28467670
040 $a MiAaPQ $c MiAaPQ
100 1 $a Mullan, Patrick. $3 3684943
245 1 0 $a Digital Cues in Multimedia Forensics = $b Digitale Spuren in der Multimedia-Forensik.
260 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2021
300 $a 137 p.
500 $a Source: Dissertations Abstracts International, Volume: 83-02, Section: B.
500 $a Advisor: Ries, Christian;Geradts, Zeno J. M. H.;Freiling, Felix C.
502 $a Thesis (Ph.D.)--Friedrich-Alexander-Universitaet Erlangen-Nuernberg (Germany), 2021.
506 $a This item must not be sold to any third party vendors.
520 $a This work presents several contributions to the scientific field of digital multimedia forensics. The field addresses questions of authentication examination and source identification of multimedia files. All investigated multimedia entities are visual information carriers, i.e., images and videos. The subfield of authentication examination proposes methods which allow to verify or falsify the authenticity of the probed media object. Questions concerning source identification help to identify the source of the media file, where the term source offers different levels of granularity, for instance, the make or the model of a camera. Finding answers to questions of this kind, requires to operate with characteristic cues which suit the investigated context. Such cues, or features, can be derived from the byte stream of the file as well as from the visual content, i.e., the decoded pixels. The selected feature domain depends on the exact problem statement. Analysis on pixel data is especially revealing if the manipulation is of good quality. In contrast, applications that work with encoded files can be efficiently deployed to work on large scales of media file. Oftentimes, file-based methods are also helpful if the quality of the visual content makes working with pixel-based applications rather difficult.Therefore, this dissertation subsumes multiple of our published works from the field of digital multimedia forensics, which propose various new developed approaches for both authentication examination and source identification. Additionally, the suggested approaches work in different domains. Through this diversity, we are capable to cover a broad field of research questions and can solve different types of challenges.The first part of this work focuses on source identification. We first illustrate that the current literature does not cover developments introduced by the omnipresent usage of smartphones as imaging devices. Smartphones work with exchangeable software imaging routines. Hence, we suggest to introduce a software components axis orthogonal to a hardware components axis, i.e., proposing a new level of source identification. With the help of a large scale study on Apple devices, we demonstrate that different software versions have characteristic properties. We use this finding to predict software versions by resorting to machine learning. The influence of software when taking the image and when postprocessing is also investigated. Further, we illustrate how images from unknown models can be automatically associated with the make of the source device.The second part of the dissertation focuses on videos and their representation in modern video codecs. We discuss neural networks in terms of their ability to identify double compression in videos. Double compression is a necessary step that happens automatically when manipulating a video. The introduced method is likewise capable to predict a specific parameter, the so-called ``group of pictures'' length in the first compression of a double compressed video. We further introduce methods which are capable to locate manipulations in pixel-domain, for example, persons spliced into the footage. We demonstrate that the same method can be successfully used to probe if video sequences originate from the same video. The methods are benchmarked against state-of-the art solutions and consistently outperform them.
520 $a Die vorliegende Arbeit leistet mehrere Beitrage zum Forschungsfeld der digitalen Multimedia-Forensik. Dieses Forschungsfeld befasst sich mit Fragestellung zur Authentizitatsprufung und zur Quellenidentifikation von Multimediadateien. Alle in dieser Dissertation untersuchten Multimediadateien sind visuelle Informationstrager, d.h. es werden Bilder und Videos untersucht. Das Teilfeld der Authentizitatsprufung stellt Methoden zur Verfugung, mit denen die Echtheit des untersuchten Medienobjekts verifiziert oder falsifiziert werden kann. Bei Fragestellungen zur Quelleninformation wird auf den Ursprung der vorliegenden Multimediadatei geschlossen, wobei es unterschiedlich scharfe Abstufungen von Ursprung, z.B. den Markenname oder den Modellname, gibt. Um diese Fragestellungen erfolgreich zu beantworten, muss im jeweiligem Kontext mit charakteristischen Spuren gearbeitet werden. Diese charakteristischen Spuren, auch Merkmale genannt, konnen sowohl aus dem kodierten Bytestrom der Datei gewonnen werden, als auch aus dem visuellen Inhalt, also den Pixeln. Die gewahlte Merkmalsdomane hangt auch von der exakten Problemstellung ab. Analysen der Pixeldaten sind insbesondere dann aufschlussreich, wenn eine Manipulation in guter Qualitat vorliegt. Im Gegensatz dazu konnen Anwendungen, die mit kodierten Dateien arbeiten, effizient grose Mengen an Mediendateien untersuchen und oftmals auch eingesetzt werden, wenn die Qualitat eine pixel-basierte Anwendung eher schwierig gestaltet.Diese Dissertation fasst deshalb mehrere von uns publizierte Arbeiten aus dem Feld der digitalen Multimedia-Forensik zusammen, die verschiedene neu entwickelte Verfahren sowohl aus der Quellenidentifikation als auch aus der Authentizitatsprufung vorstellen. Diese Verfahren arbeiten wiederum in unterschiedlichen Domanen. Durch diese Diversitat konnen wir ein breites Feld an Fragestellungen abdecken und unterschiedliche Losungsstrategien vorstellen.Der erste Teil der Arbeit befasst sich mit der Quellenidentifikation. Wir zeigen zuerst auf, dass die bereits bestehende Literatur Entwicklungen, die durch die Verwendung des Smartphones entstanden sind, insbesondere veranderbaren Bildspeicherroutinen, nicht abdeckt. Wir schlagen deshalb vor, neben einer Quellidentifikation entlang einer Hardwareachse zusatzlich eine dazu orthogonale Achse fur austauschbare Softwarekomponenten zu verwenden. Des Weiteren zeigen wir anhand einer gro"sen Studie von Bildern aus Apple-Geraten auf, dass sich unterschiedliche Softwareversionen charakteristisch unterscheiden. Wir nutzen diese Erkenntnis um mit Hilfe von maschinellem Lernen Softwareversionen zu pradizieren. Die Entwicklung und Einfluss von Software bei der Aufnahme von Bildern auf modernen Kameras sowie der Einfluss von Software auf den Aufbau der Bilddatei wird ebenfalls untersucht. Wir zeigen dabei auch auf wie Bilder von unbekannten Modellen einer bekannten Marke automatisiert zugeordnet werden konnen.Der zweite Teil der Arbeit befasst sich mit Videos und deren Reprasentation in modernen Videocodecs. Es werden neuronale Netze hinsichtlich ihrer Fahigkeit, Mehrfachkompressionen zu ermitteln, untersucht. Die Mehrfachkompression ist ein Schritt der bei einer Manipulation zwangslaufig anfallt. Die vorgestellte Methodik ist auch anwendbar, um einen bestimmten Parameter, die sogenannte "Group of Pictures" Lange der ersten Kompression in einem zweifach komprimierten Video, abzuschatzen. Des Weiteren werden pixel-basierte Methoden vorgestellt, um Manipulationen, etwa das Einfugen einer Person in ein Video, detektieren zu konnen. Es wird gezeigt, dass sich dieselbe Methodik einsetzen lasst, um zu prufen, ob Videosequenzen aus dem gleichen Video stammen. Die Verfahren werden gegen aktuelle Losungen verglichen und erziehlen durchweg bessere Ergebnisse.
590 $a School code: 0575.
650 4 $a Software. $2 gtt. $3 619355
650 4 $a Photographs. $3 627415
650 4 $a Forensic sciences. $3 606870
650 4 $a Newspapers. $3 601320
650 4 $a Video recordings. $3 575241
650 4 $a Identification. $3 827285
650 4 $a Artificial intelligence. $3 516317
650 4 $a Photography. $3 518339
650 4 $a Digitization. $3 3681264
650 4 $a False information. $3 3564362
650 4 $a Criminal investigations. $3 3564842
650 4 $a Localization. $3 3560711
650 4 $a Computer forensics. $3 3541388
650 4 $a Multimedia communications. $3 590562
650 4 $a Law enforcement. $3 607408
650 4 $a Film studies. $3 2122736
650 4 $a Mass communications. $3 3422380
650 4 $a Communication. $3 524709
650 4 $a Dissertations & theses. $3 3560115
650 4 $a Documentary films. $3 540573
650 4 $a Multimedia. $2 gtt $3 1000510
650 4 $a Classification. $3 595585
650 4 $a Law. $3 600858
650 4 $a Painting. $3 524049
690 $a 0800
690 $a 0558
690 $a 0900
690 $a 0206
690 $a 0708
690 $a 0398
690 $a 0459
710 2 $a Friedrich-Alexander-Universitaet Erlangen-Nuernberg (Germany). $3 3564652
773 0 $t Dissertations Abstracts International $g 83-02B.
790 $a 0575
791 $a Ph.D.
792 $a 2021
793 $a English
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=28467670