Current studies against DeepFake attacks are mostly passive methods that detect specific defects of DeepFake algorithms, lacking generalization ability. Meanwhile, existing active defense methods only focus on defending against face attribute manipulations, and there remain enormous challenges to establishing an active and sustainable defense mechanism for face swap detection. Therefore, we propose a novel training framework called FSD-GAN (Face Swap Detection based on Generative Adversarial Network), immune to the evolution of face swap attacks. Specifically, FSD-GAN contains three modules: the data processing module, the attack module that generates fake faces only used in training, and the defense module that consists of a fingerprint generator and a fingerprint discriminator. We embed the latent noise fingerprints generated by the fingerprint generator into face images, unperceivable to attackers visually and statistically. Once an attacker uses these protected faces to perform face swap attacks, these fingerprints will be transferred from training data (protected faces) to generative models (real-world face swap models), and they also exist in generated results (swapped faces). Our discriminator can easily detect latent noise fingerprints embedded in face images, converting the problem of face swap detection to verifying if fingerprints exist in swapped face images or not. Moreover, we alternately train the attack and defense modules under an adversarial framework, making the defense module more robust. We illustrate the effectiveness and robustness of FSD-GAN through extensive experiments, demonstrating that it can confront various face images, mainstream face swap models, and JPEG compression under different qualities.

A thorough understanding of the information dissemination process in Online Social Networks (OSNs) is crucial for enhancing user behavior analysis. While recent studies usually focus on assessing the emotional intensity of individual tweets or predicting their popularity, they frequently overlook how these tweets impact sentiment trends over time. The explosive and inflammatory nature of deliberate tweets is difficult to perceive by prediction or sentiment methods. To address this gap, we propose the multi-view Information Propagation State Awareness (IPSA) model, which aims to simultaneously assess and forecast both the popularity and sentiment strength throughout the information propagation process. Our approach begins by segmenting the information propagation into distinct time windows. Within each window, the IPSA model designs an encoder module to capture multi-view influence factors from structure, content, and time series data. Specifically, the encoder module includes a graph encoder layer based on graph attention networks to represent the backbone propagation structure formed by key nodes in the reply chain. Meanwhile, the sentiment encoder layer, utilizing an attention mechanism, extracts emotional factors present in the reply chain. Besides, we introduce a residual information prediction method that enhances the model’s precision in perceiving both popularity and sentiment intensity for each time window. Our comparative experiments, conducted on two datasets and benchmarked against State-of-the-Art (SOTA) methods, demonstrate that the IPSA model excels in predicting popularity and assessing future emotional trends in information propagation.

The user-generated social media messages usually contain considerable multimodal content. Such messages are usually short and lack explicit sentiment words. However, we can understand the sentiment associated with such messages by analyzing the context, which is essential to improve the sentiment analysis performance. Unfortunately, majority of the existing studies consider the impact of contextual information based on a single data model. In this study, we propose a novel model for performing context-aware user sentiment analysis. This model involves the semantic correlation of different modalities and the effects of tweet context information. Based on our experimental results obtained using the Twitter dataset, our approach is observed to outperform the other existing methods in analysing user sentiment.