A5101562847- Cloud Computing and Resource Management
- Advanced Data Storage Technologies
- Parallel Computing and Optimization Techniques
- Caching and Content Delivery
- IoT and Edge/Fog Computing
- Security and Verification in Computing
- Software-Defined Networks and 5G
- Privacy-Preserving Technologies in Data
- Advanced Malware Detection Techniques
- 3D Shape Modeling and Analysis
- Cryptography and Data Security
- Distributed and Parallel Computing Systems
- Cloud Data Security Solutions
- Adversarial Robustness in Machine Learning
- Advanced Memory and Neural Computing
- Software System Performance and Reliability
- Computer Graphics and Visualization Techniques
- Human Pose and Action Recognition
- Advanced Computational Techniques and Applications
- Luminescence Properties of Advanced Materials
- Generative Adversarial Networks and Image Synthesis
- Interconnection Networks and Systems
- Speech and Audio Processing
- Human Motion and Animation
- Embedded Systems Design Techniques
Zhejiang University of Science and Technology
2013-2025
Zhejiang University
2016-2025
Chongqing University of Posts and Telecommunications
2025
Durham University
2024
University of Shanghai for Science and Technology
2024
Jilin University
2022
Simon Fraser University
2006-2020
Shanghai Jiao Tong University
2017
Beijing University of Posts and Telecommunications
2009-2016
Xiamen University
2015
Vehicular ad-hoc networks (VANETs) have drawn much attention of researchers. The vehicles in VANETs frequently join and leave the networks, therefore restructure network dynamically automatically. Forwarded messages vehicular are primarily multimedia data, including structured plain text, sound, video, which require access control with efficient privacy preservation. Ciphertext-policy attribute-based encryption (CP-ABE) is adopted to meet requirements. However, solutions based on traditional...
In recent years researchers have made noticeable progresses in mesh denoising, that is, recovering high-quality 3D models from meshes corrupted with noise (raw or synthetic). Nevertheless, these state of the art approaches still fall short for robustly handling various noisy models. The main technical challenge robust denoising is to remove while maximally preserving geometric features. particular, this issue becomes more difficult considerable amount noise. paper we present a novel scheme...
Point set filtering, which aims at reconstructing noise-free point sets from their corresponding noisy inputs, is a fundamental problem in 3D geometry processing. The main challenge of filtering to preserve geometric features the underlying while same time removing noise. State-of-the-art methods still struggle with this issue: some are not designed recover sharp features, and others cannot well especially fine-scale features. In paper, we propose novel approach for robust feature-preserving...
A large amount of valuable information resides in a decentralized attributed social graph, where each user locally maintains limited view the graph. However, there exists conflicting requirement between publishing an graph and protecting privacy sensitive contained user's local data. In this paper, we aim to collect generate graphs manner while providing differential (LDP) for collected Existing LDP-based synthetic generation methods either fail preserve important properties (such as...
Emerging cloud-native distributed databases rely on local NVMe SSDs to provide high-performance and highavailable data services many cloud applications. However, the database clusters suffer from low utilization of storage because imbalance between CPU capacities within each node. For instance, OceanBase cluster, with hundreds PB capacity, only utilizes around 40% its storage. Although disaggregated (EBS) can enhance by provisioning independently demand, they performance bottlenecks high...
Model merging is a widespread technology in large language models (LLMs) that integrates multiple task-specific LLMs into unified one, enabling the merged model to inherit specialized capabilities of these LLMs. Most are sourced from open-source communities and have not undergone rigorous auditing, potentially imposing risks merging. This paper highlights an overlooked privacy risk: \textit{an unsafe could compromise other involved merging.} Specifically, we propose PhiMM, attack approach...
Large Language Models (LLMs) pose significant privacy risks, potentially leaking training data due to implicit memorization. Existing attacks primarily focus on membership inference (MIAs) or extraction attacks, but reconstructing specific personally identifiable information (PII) in LLM's remains challenging. In this paper, we propose R.R. (Recollect and Rank), a novel two-step stealing attack that enables attackers reconstruct PII entities from scrubbed where the have been masked. first...
Mobile virtualization has emerged fairly recently and is considered a valuable way to mitigate security risks on Android devices. However, major challenges in mobile include runtime, hardware, resource overhead, compatibility. In this paper, we propose lightweight solution named Condroid, which based container technology. Condroid utilizes isolation namespaces feature control cgroups feature. By leveraging them, can host multiple independent virtual machines single kernel support mutilple...
Prompt-tuning has emerged as an attractive paradigm for deploying large-scale language models due to its strong downstream task performance and efficient multitask serving ability. Despite wide adoption, we empirically show that prompt-tuning is vulnerable task-agnostic backdoors, which reside in the pretrained can affect arbitrary tasks. The state-of-the-art backdoor detection approaches cannot defend against backdoors since they hardly converge reversing triggers. To address this issue,...
In recent years, DeepFake technologies have seen widespread adoption in various domains, including entertainment and film production. However, they also been maliciously employed for disseminating false information engaging video fraud. Existing detection methods often experience significant performance degradation when confronted with unknown forgeries or exhibit limitations dealing low-quality images. To address this challenge, we introduce <italic...
Ba5-x(BO3)2(B2O5):xSm3+(0.01≤ x ≤ 0.15) orange-red phosphors were synthesized by conventional solid state reaction method. Its crystal structure, luminescence properties and concentration quenching mechanism investigated. This phosphor can be effectively excited 403 nm near ultraviolet light with the emission peaks located at 563, 600 648 nm, features a satisfactory performance. It is shown that optimal of Sm3+ ions in Ba5(BO3)2(B2O5):Sm3+ 7 mol%. The charge compensated behaviors, where Li+,...