[HTML][HTML] Continual learning for adaptive social network identification

S Magistri, D Baracchi, D Shullani, AD Bagdanov… - Pattern Recognition …, 2024 - Elsevier
The popularity of social networks as primary mediums for sharing visual content has made it
crucial for forensic experts to identify the original platform of multimedia content. Various …

[HTML][HTML] Uncovering the authorship: Linking media content to social user profiles

D Baracchi, D Shullani, M Iuliani, D Giani… - Pattern Recognition …, 2024 - Elsevier
The extensive spread of fake news on social networks is carried out by a diverse range of
users, encompassing private individuals, newspapers, and organizations. With widely …

How green is continual learning, really? Analyzing the energy consumption in continual training of vision foundation models

T Trinci, S Magistri, R Verdecchia… - arXiv preprint arXiv …, 2024 - arxiv.org
With the ever-growing adoption of AI, its impact on the environment is no longer negligible.
Despite the potential that continual learning could have towards Green AI, its environmental …

FloreView: an image and video dataset for forensic analysis

D Baracchi, D Shullani, M Iuliani, A Piva - IEEE Access, 2023 - ieeexplore.ieee.org
Linking a digital image or video to its originating device, or checking the content integrity still
represent challenging forensic tasks. Even though several technologies based on metadata …

Tiny Autoencoders are Effective Few-Shot Generative Model Detectors

L Bindini, G Bertazzini, D Baracchi… - … and Security (WIFS), 2024 - ieeexplore.ieee.org
The development of generative AI techniques such as Generative Adversarial Networks and
Diffusion Models has made it accessible to create images, often extremely realistic, that do …

Structure Matters: Analyzing Videos Via Graph Neural Networks for Social Media Platform Attribution

A Gemelli, D Shullani, D Baracchi… - ICASSP 2024-2024 …, 2024 - ieeexplore.ieee.org
Detecting the origin of a digital video within a social network is a critical task that aids law
enforcement and intelligence agencies in identifying the creators of misleading visual …

A Codec-Based Approach for Video Life-Cycle Characterization in Social Networks

G Bertazzini, D Baracchi, D Shullani… - ICASSP 2024-2024 …, 2024 - ieeexplore.ieee.org
Over the past decade, the proliferation of social networks introduced new challenges in the
multimedia forensic field, such as the identification of the originating platform. Significant …

Efficient Architectures and Incremental Methodologies for Deep Learning

S Magistri, AD Bagdanov - 2024 - flore.unifi.it
Deep Neural Networks (DNNs) are fundamental to Artificial Intelligence (AI) applications,
demonstrating remarkable performance across a diverse range of tasks in industries such as …