Multimedia data manipulation and forgery has never been easier than today, thanks to the power of Artificial Intelligence (AI). AI-generated fake content, commonly called Deepfakes, have been raising new issues and concerns, but also new challenges for the research community. The Deepfake detection task has become widely addressed, but unfortunately, approaches in the literature suffer from generalization issues. In this paper, the Face Deepfake Detection and Reconstruction Challenge is described. Two different tasks were proposed to the participants: (i) creating a Deepfake detector capable of working in an “in the wild” scenario; (ii) creating a method capable of reconstructing original images from Deepfakes. Real images from CelebA and FFHQ and Deepfake images created by StarGAN, StarGAN-v2, StyleGAN, StyleGAN2, AttGAN and GDWCT were collected for the competition. The winning teams were chosen with respect to the highest classification accuracy value (Task I) and “minimum average distance to Manhattan” (Task II). Deep Learning algorithms, particularly those based on the EfficientNet architecture, achieved the best results in Task I. No winners were proclaimed for Task II. A detailed discussion of teams’ proposed methods with corresponding ranking is presented in this paper.

Deepfake Generation Techniques are evolving at a rapid pace, making it possible to create realistic manipulated images and videos and endangering the serenity of modern society. The continual emergence of new and varied techniques brings with it a further problem to be faced, namely the ability of deepfake detection models to update themselves promptly in order to be able to identify manipulations carried out using even the most recent methods. This is an extremely complex problem to solve, as training a model requires large amounts of data, which are difficult to obtain if the deepfake generation method is too recent. Moreover, continuously retraining a network would be unfeasible. In this paper, we ask ourselves if, among the various deep learning techniques, there is one that is able to generalise the concept of deepfake to such an extent that it does not remain tied to one or more specific deepfake generation methods used in the training set. We compared a Vision Transformer with an EfficientNetV2 on a cross-forgery context based on the ForgeryNet dataset. From our experiments, It emerges that EfficientNetV2 has a greater tendency to specialize often obtaining better results on training methods while Vision Transformers exhibit a superior generalization ability that makes them more competent even on images generated with new methodologies.

In this paper, we present MINTIME, a video deepfake detection method that effectively captures spatial and temporal inconsistencies in videos that depict multiple individuals and varying face sizes. Unlike previous approaches that either employ simplistic a-posteriori aggregation schemes, i.e., averaging or max operations, or only focus on the largest face in the video, our proposed method learns to accurately detect spatio-temporal inconsistencies across multiple identities in a video through a Spatio-Temporal Transformer combined with a Convolutional Neural Network backbone. This is achieved through an Identity-aware Attention mechanism that applies a masking operation on the face sequence to process each identity independently, which enables effective video-level aggregation. Furthermore, our system incorporates two novel embedding schemes: (i) the Temporal Coherent Positional Embedding, which encodes the temporal information of the face sequences of each identity, and (ii) the Size Embedding, which captures the relative sizes of the faces to the video frames. MINTIME achieves state-of-the-art performance on the ForgeryNet dataset, with a remarkable improvement of up to 14% AUC in videos containing multiple people. Moreover, it demonstrates very robust generalization capabilities in cross-forgery and cross-dataset settings. The code is publicly available at: https://github.com/davide-coccomini/MINTIME-Multi-Ident ity-size-iNvariant-TIMEsformer-for-Video-Deepfake-Detection.

The increasing use of deep learning techniques to manipulate images and videos, commonly referred to as “deepfakes”, is making it more challenging to differentiate between real and fake content, while various deepfake detection systems have been developed, they often struggle to detect deepfakes in real-world situations. In particular, these methods are often unable to effectively distinguish images or videos when these are modified using novel techniques which have not been used in the training set. In this study, we carry out an analysis of different deep learning architectures in an attempt to understand which is more capable of better generalizing the concept of deepfake. According to our results, it appears that Convolutional Neural Networks (CNNs) seem to be more capable of storing specific anomalies and thus excel in cases of datasets with a limited number of elements and manipulation methodologies. The Vision Transformer, conversely, is more effective when trained with more varied datasets, achieving more outstanding generalization capabilities than the other methods analysed. Finally, the Swin Transformer appears to be a good alternative for using an attention-based method in a more limited data regime and performs very well in cross-dataset scenarios. All the analysed architectures seem to have a different way to look at deepfakes, but since in a real-world environment the generalization capability is essential, based on the experiments carried out, the attention-based architectures seem to provide superior performances.

In recent years, the field of artificial intelligence has witnessed a remarkable surge in the generation of synthetic images, driven by advancements in deep learning techniques. These synthetic images, often created through complex algorithms, closely mimic real photographs, blurring the lines between reality and artificiality. This proliferation of synthetic visuals presents a pressing challenge: how to accurately and reliably distinguish between genuine and generated images. This article, in particular, explores the task of detecting images generated by text-to-image diffusion models, highlighting the challenges and peculiarities of this field. To evaluate this, we consider images generated from captions in the MSCOCO and Wikimedia datasets using two state-of-the-art models: Stable Diffusion and GLIDE. Our experiments show that it is possible to detect the generated images using simple multi-layer perceptrons (MLPs), starting from features extracted by CLIP or RoBERTa, or using traditional convolutional neural networks (CNNs). These latter models achieve remarkable performances in particular when pretrained on large datasets. We also observe that models trained on images generated by Stable Diffusion can occasionally detect images generated by GLIDE, but only on the MSCOCO dataset. However, the reverse is not true. Lastly, we find that incorporating the associated textual information with the images in some cases can lead to a better generalization capability, especially if textual features are closely related to visual ones. We also discovered that the type of subject depicted in the image can significantly impact performance. This work provides insights into the feasibility of detecting generated images and has implications for security and privacy concerns in real-world applications. The code to reproduce our results is available at:https://github.com/davide-coccomini/Detecting-Images-Generated-by-Diffusers.