國立虎尾科技大學 |

Strengthening Image Generative AI: Integrating Fingerprinting and Revision Methods for Enhanced Safety and Control /

Record Type:	Language materials, printed : Monograph/item
Title/Author:	Strengthening Image Generative AI: Integrating Fingerprinting and Revision Methods for Enhanced Safety and Control // Changhoon Kim.
Author:	Kim, Changhoon,
Description:	1 electronic resource (130 pages)
Notes:	Source: Dissertations Abstracts International, Volume: 86-01, Section: B.
Contained By:	Dissertations Abstracts International86-01B.
Subject:	Computer engineering. -
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31334701
ISBN:	9798383189443

Strengthening Image Generative AI: Integrating Fingerprinting and Revision Methods for Enhanced Safety and Control /
Kim, Changhoon,

Strengthening Image Generative AI: Integrating Fingerprinting and Revision Methods for Enhanced Safety and Control /Changhoon Kim. - 1 electronic resource (130 pages)

Source: Dissertations Abstracts International, Volume: 86-01, Section: B.

In the rapidly evolving field of Generative Artificial Intelligence (Gen-AI) for imaging, models such as DALL·E3 and Stable Diffusion have transitioned from theoretical concepts to practical tools with significant impact across various sectors including entertainment, art, journalism, and education. These advancements represent a substantial technological evolution, enhancing creative and professional practices. However, the widespread accessibility of Gen-AI also facilitates misuse by malicious actors who create deepfakes and spread misinformation, posing serious risks to societal well-being and privacy. This dissertation addresses these critical challenges by focusing on enhancing the reliability of Image GenAI models through the identification and mitigation of inherent vulnerabilities. It presents the development of computational tools and frameworks aimed at enabling better community oversight, ensuring the secure and responsible use of Gen-AI. Key contributions of this work include the introduction of innovative fingerprinting techniques that trace malicious Gen-AI outputs back to their sources, and the implementation of strategies to prevent the generation of unauthorized content. These efforts collectively strengthen the robustness and accountability of Gen-AI technologies, particularly in sensitive applications.

English

ISBN: 9798383189443Subjects--Topical Terms:

569006
Computer engineering.
Subjects--Index Terms:

Reliable AI

Strengthening Image Generative AI: Integrating Fingerprinting and Revision Methods for Enhanced Safety and Control /
LDR:02854nam a22004333i 4500 001 1157833
005 20250603111424.5
006 m o d
007 cr|nu||||||||
008 250804s2024 miu||||||m |||||||eng d
020 $a 9798383189443
035 $a (MiAaPQD)AAI31334701
035 $a AAI31334701
040 $a MiAaPQD $b eng $c MiAaPQD $e rda
100 1 $a Kim, Changhoon, $e author. $3 1484110
245 1 0 $a Strengthening Image Generative AI: Integrating Fingerprinting and Revision Methods for Enhanced Safety and Control / $c Changhoon Kim.
264 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2024
300 $a 1 electronic resource (130 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertations Abstracts International, Volume: 86-01, Section: B.
500 $a Advisors: Yang, Yezhou Committee members: Xue, Guoliang; Pedrielli, Giulia; Lee, Kookjin.
502 $b Ph.D. $c Arizona State University $d 2024.
520 $a In the rapidly evolving field of Generative Artificial Intelligence (Gen-AI) for imaging, models such as DALL·E3 and Stable Diffusion have transitioned from theoretical concepts to practical tools with significant impact across various sectors including entertainment, art, journalism, and education. These advancements represent a substantial technological evolution, enhancing creative and professional practices. However, the widespread accessibility of Gen-AI also facilitates misuse by malicious actors who create deepfakes and spread misinformation, posing serious risks to societal well-being and privacy. This dissertation addresses these critical challenges by focusing on enhancing the reliability of Image GenAI models through the identification and mitigation of inherent vulnerabilities. It presents the development of computational tools and frameworks aimed at enabling better community oversight, ensuring the secure and responsible use of Gen-AI. Key contributions of this work include the introduction of innovative fingerprinting techniques that trace malicious Gen-AI outputs back to their sources, and the implementation of strategies to prevent the generation of unauthorized content. These efforts collectively strengthen the robustness and accountability of Gen-AI technologies, particularly in sensitive applications.
546 $a English
590 $a School code: 0010
650 4 $a Computer engineering. $3 569006
650 4 $a Information technology. $3 559429
653 $a Reliable AI
653 $a Secure AI
653 $a Generative Artificial Intelligence
653 $a Entertainment
653 $a Fingerprinting
690 $a 0464
690 $a 0489
690 $a 0800
710 2 $a Arizona State University. $b Computer Engineering. $3 1184417
720 1 $a Yang, Yezhou $e degree supervisor.
773 0 $t Dissertations Abstracts International $g 86-01B.
790 $a 0010
791 $a Ph.D.
792 $a 2024
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=31334701