The paper investigates the security of a selective coordinate encryption scheme for point clouds used in volumetric video. By training PointNet and Random Forest models on partially encrypted data, the authors show that fully encrypting all X coordinates is difficult to reconstruct, whereas encrypting every second X coordinate allows accurate recovery of the hidden values. The study highlights that the encryption granularity critically determines the vulnerability of the scheme.
Vision Transformers (ViTs) are increasingly used in split-inference systems, where edge devices transmit intermediate token representations to a remote cloud. In this setting, token reduction lowers c...
arXiv:2409. 01062v4 Announce Type: replace Abstract: Model Inversion (MI) attacks pose a significant privacy threat by reconstructing private training data from machine learning models.
The paper investigates privacy risks in Vision Transformer (ViT) split‑inference systems that use token reduction and token shuffling to lower computation and communication costs. It shows that even after token shuffling, transmitted token embeddings still contain enough positional information for a new attack, the Spatially Aligned Reconstruction Attack (SARA), which predicts token positions, restores spatial layout, fills missing embeddings with a masked autoencoder, and reconstructs the input image. While token reduction offers stronger protection, significant leakage remains when retained tokens preserve semantic and positional cues, and the authors propose a lightweight edge‑side defense that removes positional embeddings and adapts transformer blocks via knowledge distillation to reduce SARA’s effectiveness without harming downstream accuracy.
arXiv:2610.35729v1 Announce Type: cross Abstract: Imagery captured by unmanned aerial vehicle (UAV) systems increasingly raises privacy concerns, as UAV aerial surveillance routinely captures sensiti...
arXiv:2608.11645v2 Announce Type: replace Abstract: Volumetric video streaming turns privacy into a 3D, multi-view problem. Unlike ordinary video, where sensitive content can often be redacted frame...
arXiv:2609.13823v1 Announce Type: new Abstract: Point cloud data face serious semantic privacy risks during acquisition, transmission, and cross-institutional sharing. Existing methods mostly rely on...
arXiv:2607. 08867v1 Announce Type: cross Abstract: Cloud-based deep learning enables large-scale medical image analysis but raises significant privacy concerns when sensitive patient images are outsourced for model development.
The paper presents a system for real‑time mixed reality that reduces bandwidth and encryption overhead by downsampling point cloud data at the server and partially encrypting it. The client decrypts the data and uses a machine‑learning super‑resolution model to reconstruct the full‑resolution point clouds. Experiments show a nearly linear decrease in bandwidth and latency, with the super‑resolution model achieving low reconstruction error and modest inference time.
The unprecedented growth of computer vision applications, such as surveillance systems and social media, raises security and visual privacy concerns, especially when data is stored on cloud servers. Image obfuscation offers a way to preserve visual privacy while maintaining an adequate level of usability; thus, it has been a topic of great interest in recent years.
ComplicitSplat is a novel black‑box attack that leverages 3D Gaussian Splatting (3DGS) shading to create viewpoint‑specific camouflage, embedding adversarial content into scene objects that is only visible from certain angles. The method does not require access to model architecture or weights and can successfully fool a range of popular object detectors—including single‑stage, multi‑stage, and transformer‑based models—on both real‑world physical objects and synthetic scenes. This demonstrates that downstream models using 3DGS are vulnerable to adversarial manipulation.
arXiv:2412. 12640v2 Announce Type: replace Abstract: The increasing demand for data privacy, alongside the benefits of aggregating data from networked devices, has catalyzed the emergence of federated learning (FL).
arXiv:2508.19819v3 Announce Type: replace-cross Abstract: Gradient inversion attacks are often presented as a serious privacy threat in federated learning, with recent work reporting increasingly str...