Daniele Baracchi

Francesca Lanzillotta, Chiara Albisani, Davide Pucci et al.

In many learning tasks, certain requirements on the processing of individual data samples should arguably be formalized as strict constraints in the underlying optimization problem, rather than by means of arbitrary penalties. We show that, in these scenarios, learning can be carried out exploiting a sequential penalty method that allows to properly deal with constraints. The proposed algorithm is shown to possess convergence guarantees under assumptions that are reasonable in deep learning scenarios. Moreover, the results of experiments on image processing tasks show that the method is indeed viable to be used in practice.

Giulia Bertazzini, Chiara Albisani, Daniele Baracchi et al.

With the growing adoption of AI image generation, in conjunction with the ever-increasing environmental resources demanded by AI, we are urged to answer a fundamental question: What is the environmental impact hidden behind each image we generate? In this research, we present a comprehensive empirical experiment designed to assess the energy consumption of AI image generation. Our experiment compares 17 state-of-the-art image generation models by considering multiple factors that could affect their energy consumption, such as model quantization, image resolution, and prompt length. Additionally, we consider established image quality metrics to study potential trade-offs between energy consumption and generated image quality. Results show that image generation models vary drastically in terms of the energy they consume, with up to a 46x difference. Image resolution affects energy consumption inconsistently, ranging from a 1.3x to 4.7x increase when doubling resolution. U-Net-based models tend to consume less than Transformer-based one. Model quantization instead results to deteriorate the energy efficiency of most models, while prompt length and content have no statistically significant impact. Improving image quality does not always come at the cost of a higher energy consumption, with some of the models producing the highest quality images also being among the most energy efficient ones.

Andrea Montibeller, Dasara Shullani, Daniele Baracchi et al.

The growing presence of AI-generated videos on social networks poses new challenges for deepfake detection, as detectors trained under controlled conditions often fail to generalize to real-world scenarios. A key factor behind this gap is the aggressive, proprietary compression applied by platforms like YouTube and Facebook, which launder low-level forensic cues. However, replicating these transformations at scale is difficult due to API limitations and data-sharing constraints. For these reasons, we propose a first framework that emulates the video sharing pipelines of social networks by estimating compression and resizing parameters from a small set of uploaded videos. These parameters enable a local emulator capable of reproducing platform-specific artifacts on large datasets without direct API access. Experiments on FaceForensics++ videos shared via social networks demonstrate that our emulated data closely matches the degradation patterns of real uploads. Furthermore, detectors fine-tuned on emulated videos achieve comparable performance to those trained on actual shared media. Our approach offers a scalable and practical solution for bridging the gap between lab-based training and real-world deployment of deepfake detectors, particularly in the underexplored domain of compressed video content.

Pengpeng Yang, Daniele Baracchi, Massimo Iuliani et al.

Most video forensic techniques look for traces within the data stream that are, however, mostly ineffective when dealing with strongly compressed or low resolution videos. Recent research highlighted that useful forensic traces are also left in the video container structure, thus offering the opportunity to understand the life-cycle of a video file without looking at the media stream itself. In this paper we introduce a container-based method to identify the software used to perform a video manipulation and, in most cases, the operating system of the source device. As opposed to the state of the art, the proposed method is both efficient and effective and can also provide a simple explanation for its decisions. This is achieved by using a decision-tree-based classifier applied to a vectorial representation of the video container structure. We conducted an extensive validation on a dataset of 7000 video files including both software manipulated contents (ffmpeg, Exiftool, Adobe Premiere, Avidemux, and Kdenlive), and videos exchanged through social media platforms (Facebook, TikTok, Weibo and YouTube). This dataset has been made available to the research community. The proposed method achieves an accuracy of 97.6% in distinguishing pristine from tampered videos and classifying the editing software, even when the video is cut without re-encoding or when it is downscaled to the size of a thumbnail. Furthermore, it is capable of correctly identifying the operating system of the source device for most of the tampered videos.

Francesco Orsini, Daniele Baracchi, Paolo Frasconi

We introduce an architecture based on deep hierarchical decompositions to learn effective representations of large graphs. Our framework extends classic R-decompositions used in kernel methods, enabling nested part-of-part relations. Unlike recursive neural networks, which unroll a template on input graphs directly, we unroll a neural network template over the decomposition hierarchy, allowing us to deal with the high degree variability that typically characterize social network graphs. Deep hierarchical decompositions are also amenable to domain compression, a technique that reduces both space and time complexity by exploiting symmetries. We show empirically that our approach is able to outperform current state-of-the-art graph classification methods on large social network datasets, while at the same time being competitive on small chemobiological benchmark datasets.