A5043619812- Privacy-Preserving Technologies in Data
- Cryptography and Data Security
- Adversarial Robustness in Machine Learning
- Stochastic Gradient Optimization Techniques
- Speech and Audio Processing
- Bluetooth and Wireless Communication Technologies
- Advanced Steganography and Watermarking Techniques
- Internet Traffic Analysis and Secure E-voting
- Face recognition and analysis
- Indoor and Outdoor Localization Technologies
- Chaos-based Image/Signal Encryption
- Generative Adversarial Networks and Image Synthesis
- Data Quality and Management
- Network Security and Intrusion Detection
- Algorithms and Data Compression
Xidian University
2021-2024
Henan University of Technology
2018
Pattern matching is widely used in applications such as genomic data query analysis, network intrusion detection, and deep packet inspection (DPI). Performing pattern on plaintext straightforward, but the need to protect security of analyzed patterns can significantly complicate process. Due privacy issues patterns, researchers begin explore encrypted data. However, existing solutions are typically built static methods, lacking dynamism, namely, inability perform addition or deletion...
In this paper, we present a non-destructive and economic wheat moisture detection system with commodity WiFi. First, experimentally validate the feasibility of by using CSI amplitude phase difference data. We then design Wi-Wheat system, where data preprocessing, feature extraction support vector machine (SVM) classification are implemented for processing module. For employ outlier detection, normalization eliminating noise obtaining clear Then, consider principal component analysis (PCA)...
Federated learning (FL), a cooperative distributed framework, has been employed in various intelligent Internet of Things (IoT) applications (e.g., smart health-care, home, and industry). However, there may be malicious devices these IoT inferring other devices' privacy or destroying the uploaded model parameters. Besides, due to heterogeneity devices, it is difficult for existing synchronized FL effectively train models through non-identical independently (non-IID) local data sets. To...
Federated learning (FL) in Internet of Things (IoT) applications facilitates the collaborative training a global model across distributed devices with server. Despite its potential, nature and vulnerability IoT render FL susceptible to Byzantine attacks. Existing approaches counter these attacks are often impractical real-world scenarios, mainly due challenges posed by nonindependent identically (non-IID) data high-dimensional common devices. To address challenges, we propose Guard-FL, an...
Outsourcing convolutional neural network (CNN) inference services to the cloud is extremely beneficial, yet raises critical privacy concerns on proprietary model parameters of provider and private input data user. Previous studies have indicated that some cryptographic tools such as secure multi-party computation (MPC) can be used achieve outsourced inferences. However, MPC-based approaches often require a large number communication rounds across two or more non-colluding servers, which make...
Federated Learning, a distributed machine learning paradigm, is susceptible to Byzantine attacks since the attacker can manipulate clients' local data and models compromise performance of global model.Recently, there have been wealth server-side defenses that mitigate by removing or limiting impact malicious models.Nevertheless, easily circumvent these approaches rely solely on single defense, stemming from high dimensionality variety attacks.Therefore, we propose Basalt, an efficient...
The advances in machine learning technology has promoted its great potential for deep neural network (DNN) inference powered applications of Internet Things (IoT), such as facial verification cameras and speech recognition assistants. current deployment these also raises serious privacy concerns, especially when sensitive individual information is accessed easily by various IoT devices. Fortunately, the cryptography-based solutions are able to execute secure without infringing user’s raw...
Federated Learning (FL) shows its great vitality for terminal devices of Internet Things (IoT) in privacy protection. However, the amount data on each device IoT are imbalanced which makes global training model rarely effective. Although Generative Adversarial Network (GAN) can generate to alleviate above problem, it has characteristic instability and still carries risk breach. In this paper, we propose a novel privacy-preserving federated GAN framework, named P-FedGAN, train generative with...