3D reconstruction and simulation, although interrelated, have distinct objectives: reconstruction requires a flexible 3D representation that can adapt to diverse scenes, while simulation needs a structured representation to model motion principles effectively. This paper introduces the Mesh-adsorbed Gaussian Splatting (MaGS) method to address this challenge. MaGS constrains 3D Gaussians to roam near the mesh, creating a mutually adsorbed mesh-Gaussian 3D representation. Such representation harnesses both the rendering flexibility of 3D Gaussians and the structured property of meshes. To achieve this, we introduce RMD-Net, a network that learns motion priors from video data to refine mesh deformations, alongside RGD-Net, which models the relative displacement between the mesh and Gaussians to enhance rendering fidelity under mesh constraints. To generalize to novel, user-defined deformations beyond input video without reliance on temporal data, we propose MPE-Net, which leverages inherent mesh information to bootstrap RMD-Net and RGD-Net. Due to the universality of meshes, MaGS is compatible with various deformation priors such as ARAP, SMPL, and soft physics simulation. Extensive experiments on the D-NeRF, DG-Mesh, and PeopleSnapshot datasets demonstrate that MaGS achieves state-of-the-art performance in both reconstruction and simulation.

Microarchitectural attacks continue to present a significant security risk to computer systems, irrespective of the presence of software vulnerabilities. Particularly concerning among these are spectre-style breaches, capitalizing on speculative execution, a crucial feature of modern CPUs that is designed to optimize performance. Recently, several testing tools have been devised to detect speculative leaks within commercial (black-box) CPUs automatically.Notwithstanding the prevalence of microarchitectural leaks, the corresponding codes given by researchers to outline these vulnerabilities are often tailored to particular processor types. Verifying the existence of similar conditions in various processors typically demands manual modifications to code. Given the development of fuzz testing technology in the field of transient execution, this paper investigates the characteristics of template-based and model-based methods of fuzzing tools, subsequently proposing an automated process that based on fuzzing to minimize the manual effort demanded to test and verify vulnerabilities in differing processor types. This method can substantially reduce the time consumed when assessing if a processor is susceptible to recognized attack forms.