A closer look at gan priors: Exploiting intermediate features for enhanced model inversion attacks

Y Qiu, H Fang, H Yu, B Chen, MK Qiu… - European Conference on …, 2024 - Springer
Abstract Model Inversion (MI) attacks aim to reconstruct privacy-sensitive training data from
released models by utilizing output information, raising extensive concerns about the …

Gi-nas: Boosting gradient inversion attacks through adaptive neural architecture search

W Yu, H Fang, B Chen, X Sui, C Chen, H Wu… - arxiv preprint arxiv …, 2024 - arxiv.org
Gradient Inversion Attacks invert the transmitted gradients in Federated Learning (FL)
systems to reconstruct the sensitive data of local clients and have raised considerable …

Editable-DeepSC: Reliable Cross-Modal Semantic Communications for Facial Editing

B Chen, W Yu, Q Zhang, ST **a - arxiv preprint arxiv:2411.15702, 2024 - arxiv.org
Real-time computer vision (CV) plays a crucial role in various real-world applications, whose
performance is highly dependent on communication networks. Nonetheless, the data …

Semantic communications: the new paradigm behind beyond 5G technologies

G Fernandes, H Fontes, R Campos - arxiv preprint arxiv:2406.00754, 2024 - arxiv.org
Each generation of cellular networks is characterized by its distinct capabilities and
innovations, which reflect the significant milestones reached with each new release. 5G has …

Retrievals Can Be Detrimental: A Contrastive Backdoor Attack Paradigm on Retrieval-Augmented Diffusion Models

H Fang, X Sui, H Yu, J Kong, S Yu, B Chen… - arxiv preprint arxiv …, 2025 - arxiv.org
Diffusion models (DMs) have recently demonstrated remarkable generation capability.
However, their training generally requires huge computational resources and large-scale …