LUCA MAIANO

Dottore di ricerca

ciclo: XXXV


supervisore: Irene Amerini
co-supervisore: Aris Anagnostopoulos

Titolo della tesi: Media forensics investigations: from the origin to the authenticity of digital content

In recent years, we have observed a massive change in how information is exchanged. On the one hand, the explosion of social media has given birth to a new way of communicating and exchanging news, media, and ideas. On the other hand, the advancement of content manipulation and generation technologies have led to tools capable of recreating incredibly realistic artificial content. All this poses new challenges in verifying the authenticity and integrity of online content. Whenever we come across new media, we must understand its origin, whether it is real or deliberately modified, and verify its authenticity. In this thesis, we will analyze each problem, offering an overview of possible solutions. The first challenge to solve when encountering multimedia content is reconstructing its source. This problem is as essential for verifying online news as for forensic investigations, where an image or video can represent evidence of a crime. Given a media, we wonder if it was captured with a specific offending camera model or if it was instead downloaded from a social platform. Solving this problem means analyzing the compression traces left in the file when it is captured or uploaded to a platform. To solve this challenge, we propose to train neural networks that learn to distinguish these traces, which we define as fingerprints. Specifically, we will show how these fingerprints change from camera to camera and when content is uploaded to a social network, making it possible to reconstruct the source of origin without relying on information such as metadata that can often be modified or deleted. Another significant problem is that of verifying the authenticity of information. Recent advances in the development of artificial intelligence enable the generation of incredibly realistic content: deepfakes. On the one hand, this opens the doors to new applications in entertainment and creativity. On the other hand, it introduces a new generation of super-realistic fake content. The recognition of these contents is possible thanks to a set of factors. First, many of these techniques introduce semantic inconsistencies that are difficult to correct; furthermore, each generative technique leaves specific fingerprints similar to those left by camera models or social media. We will analyze possible strategies for recognizing fake content by exploiting these inconsistencies. All the challenges mentioned so far have one problem in common. Data and information continually evolve, making standard detectors less and less robust as time passes. This is especially true with news, which constantly evolves as events worldwide grow. To prevent this from happening, fake news detectors must continuously learn to classify new information. The last part of this thesis will be dedicated to this topic. On the one hand, we will introduce a continuous learning strategy that allows a detector to learn to classify new news as it is published. Subsequently, we will analyze the vulnerabilities of these techniques concerning a new type of adversary attack. Finally, we will discuss two forensic applications in the fields of ground to aerial matching and insurance.

Produzione scientifica

11573/1719609 - 2024 - Editorial for pattern recognition letters special issue on Advances in Disinformation Detection and Media Forensics
Amerini, I.; Sanchez, V.; Maiano, L. - 01m Editorial/Introduzione in rivista
rivista: PATTERN RECOGNITION LETTERS (Elsevier BV:PO Box 211, 1000 AE Amsterdam Netherlands:011 31 20 4853757, 011 31 20 4853642, 011 31 20 4853641, EMAIL: nlinfo-f@elsevier.nl, INTERNET: http://www.elsevier.nl, Fax: 011 31 20 4853598) pp. 21-22 - issn: 0167-8655 - wos: (0) - scopus: 2-s2.0-85203414388 (0)

11573/1710641 - 2024 - A guided-based approach for deepfake detection: RGB-depth integration via features fusion
Leporoni, G.; Maiano, L.; Papa, L.; Amerini, I. - 01a Articolo in rivista
rivista: PATTERN RECOGNITION LETTERS (Elsevier BV:PO Box 211, 1000 AE Amsterdam Netherlands:011 31 20 4853757, 011 31 20 4853642, 011 31 20 4853641, EMAIL: nlinfo-f@elsevier.nl, INTERNET: http://www.elsevier.nl, Fax: 011 31 20 4853598) pp. 99-105 - issn: 0167-8655 - wos: WOS:001228815000001 (2) - scopus: 2-s2.0-85189752625 (2)

11573/1710645 - 2024 - Human Versus Machine: A Comparative Analysis in Detecting Artificial Intelligence-Generated Images
Maiano, L.; Benova, A.; Papa, L.; Stockner, M.; Marchetti, M.; Convertino, G.; Mazzoni, G.; Amerini, I. - 01a Articolo in rivista
rivista: IEEE SECURITY & PRIVACY (IEEE Computer Society.) pp. 77-86 - issn: 1540-7993 - wos: WOS:001226314400009 (0) - scopus: 2-s2.0-85193817595 (1)

11573/1710643 - 2024 - CMDD: A novel multimodal two-stream CNN deepfakes detector
Mongelli, L.; Maiano, L.; Amerini, I. - 04b Atto di convegno in volume
congresso: 4th Workshop on Reducing Online Misinformation through Credible Information Retrieval, ROMCIR 2024 (Glasgow; UK)
libro: Proceedings of the 4th Workshop on Reducing Online Misinformation through Credible Information Retrieval co-located with the 46th European Conference on Information Retrieval (ECIR 2024) - ()

11573/1693655 - 2023 - An Automated Ground-to-Aerial Viewpoint Localization for Content Verification
Bonaventura, Tania Sari; Maiano, Luca; Papa, Lorenzo; Amerini, Irene - 04b Atto di convegno in volume
congresso: 24th International Conference on Digital Signal Processing, DSP 2023 (Rhodes; Greece)
libro: 2023 24th International Conference on Digital Signal Processing (DSP) - (979-8-3503-3959-8)

11573/1719612 - 2023 - DepthFake: A Depth-Based Strategy for Detecting Deepfake Videos
Maiano, L.; Papa, L.; Vocaj, K.; Amerini, I. - 04b Atto di convegno in volume
congresso: 26th International Conference on Pattern Recognition, ICPR 2022 (Canada)
libro: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) - (9783031377440; 9783031377457)

11573/1684452 - 2023 - A deep-learning–based antifraud system for car-insurance claims
Maiano, Luca; Montuschi, Antonio; Caserio, Marta; Ferri, Egon; Kieffer, Federico; Germanò, Chiara; Baiocco, Lorenzo; Ricciardi Celsi, Lorenzo; Amerini, Irene; Anagnostopoulos, Aris - 01a Articolo in rivista
rivista: EXPERT SYSTEMS WITH APPLICATIONS (Oxford, United Kingdom: Elsevier Science Limited) pp. - - issn: 0957-4174 - wos: WOS:001023537400001 (9) - scopus: 2-s2.0-85162016647 (11)

11573/1693691 - 2023 - On the use of Stable Diffusion for creating realistic faces: from generation to detection
Papa, Lorenzo; Faiella, Lorenzo; Corvitto, Luca; Maiano, Luca; Amerini, Irene - 04b Atto di convegno in volume
congresso: 11th International Workshop on Biometrics and Forensics, IWBF 2023 (Barcelona; Spain)
libro: 2023 11th International Workshop on Biometrics and Forensics (IWBF) - (979-8-3503-3607-8)

11573/1695496 - 2023 - INTELLIGENZA ARTIFICIALE E RICONOSCIMENTO DI VOLTI
Stockner, Mara; Marchetti, Michela; Papa, Lorenzo; Maiano, Luca; Convertino, Gianmarco; Amerini, Irene - 04d Abstract in atti di convegno
congresso: XXIX CONGRESSO Associazione Italiana di Psicologia - Sezione Sperimentale (Lucca, Italy)
libro: ATTI DEL XXIX CONGRESSO AIP - ()

11573/1652326 - 2022 - DepthFake: a depth-based strategy for detecting Deepfake videos
Maiano, Luca; Papa, Lorenzo; Vocaj, Ketbjano; Amerini, Irene - 04b Atto di convegno in volume
congresso: International Conference on Pattern Recognition 2022 (Montreal, Quebec)
libro: ICPR 2022 Workshop on Artificial Intelligence for Multimedia Forensics and Disinformation Detection - ()

11573/1622955 - 2021 - Deep learning for multimedia forensics
Amerini, I.; Anagnostopoulos, A.; Maiano, L.; Ricciardi Celsi, L. - 01a Articolo in rivista
rivista: FOUNDATIONS AND TRENDS IN COMPUTER GRAPHICS AND VISION (Hanover, MA : Now Publishers, Inc., c2005-) pp. 309-457 - issn: 1572-2740 - wos: WOS:000692549000001 (12) - scopus: 2-s2.0-85113781771 (19)

11573/1615509 - 2021 - Learning double-compression video fingerprints left from social-media platforms
Amerini, I.; Anagnostopoulos, A.; Maiano, L.; Ricciardi Celsi, Lorenzo - 04b Atto di convegno in volume
congresso: ICASSP 2021 (Toronto; Canada)
libro: ICASSP 2021 - 2021 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) - ()

11573/1622915 - 2021 - Identification of social-media platform of videos through the use of shared features
Maiano, L.; Amerini, I.; Ricciardi Celsi, L.; Anagnostopoulos, A. - 01a Articolo in rivista
rivista: JOURNAL OF IMAGING (Basel : MDPI AG, 2015-) pp. 1-16 - issn: 2313-433X - wos: WOS:000689224200001 (6) - scopus: 2-s2.0-85113719340 (14)

11573/1344478 - 2019 - Data-driven intrusion detection for ambient intelligence
Chatzigiannakis, I.; Maiano, Luca; Trakadas, P.; Anagnostopoulos, A.; Bacci, F.; Karkazis, P.; Spirakis, P. G.; Zahariadis, T. - 04b Atto di convegno in volume
congresso: 15th European Conference on Ambient Intelligence, AmI 2019 (Rome; Italy)
libro: Ambient Intelligence - (978-3-030-34254-8; 978-3-030-34255-5)

© Università degli Studi di Roma "La Sapienza" - Piazzale Aldo Moro 5, 00185 Roma