A5059620114- Anomaly Detection Techniques and Applications
- Cooperative Communication and Network Coding
- Adversarial Robustness in Machine Learning
- Water Systems and Optimization
- Mobile Ad Hoc Networks
- Full-Duplex Wireless Communications
- IoT and Edge/Fog Computing
- Time Series Analysis and Forecasting
- Imbalanced Data Classification Techniques
- Gait Recognition and Analysis
- Quantum and electron transport phenomena
- Integrated Circuits and Semiconductor Failure Analysis
- Network Security and Intrusion Detection
- Machine Learning and Data Classification
- IoT Networks and Protocols
- Advanced Software Engineering Methodologies
- Real-Time Systems Scheduling
- Software System Performance and Reliability
- COVID-19 diagnosis using AI
- Context-Aware Activity Recognition Systems
- Wireless Communication Security Techniques
- Indoor and Outdoor Localization Technologies
- Quantum Computing Algorithms and Architecture
- Wireless Networks and Protocols
- Quantum Information and Cryptography
National University of Defense Technology
2021-2022
Institute of Automation
2022
Chinese Academy of Sciences
2022
Zhejiang University of Technology
2012-2021
Deep autoencoder-based methods are the majority of deep anomaly detection. An autoencoder learning on training data is assumed to produce higher reconstruction error for anomalous samples than normal and thus can distinguish anomalies from data. However, this assumption does not always hold in practice, especially unsupervised detection, where contaminated. We observe that generalizes so well it reconstruct both well, leading poor detection performance. Besides, we find performance stable...
The considerable risk of falls and the substantial increase in elderly population make automatic fall detection system become very important. Existing systems using accelerometer as detector are often designed based on an empirical acceleration threshold to differentiate from normal activities. In this paper, we design method under Neyman-Pearson framework. An optimal can be obtained which meets specified false alarm rate while maximizing probability. We use TelosW mote with detector, is...
Autoencoder based methods are the majority of deep unsupervised outlier detection methods. However, these perform not well on complex image datasets and suffer from noise introduced by outliers, especially when ratio is high. In this paper, we propose a framework named Transformation Invariant AutoEncoder (TIAE), which can achieve stable high performance detection. First, instead using conventional autoencoder, transformation invariant autoencoder to do better representation learning for...
Existing unsupervised outlier detection (OD) solutions face a grave challenge with surging visual data like images. Although deep neural networks (DNNs) prove successful for data, OD remains difficult due to OD's nature. This paper proposes novel framework named E 3Outlier that can perform effective and end-to-end removal. Its core idea is introduce self-supervision into OD. Specifically, our major solution adopt discriminative learning paradigm creates multiple pseudo classes from given...
The network coding systems are vulnerable to pollution attacks, in which the adversary forwarders can inject polluted or forged messages into networks. In this paper, we propose a homomorphic signature scheme provide security protection against mainly uses property of based on Paillier cryptosystem. On one hand, it guarantees verify received effectively. If verification is failed, message assumed be and should discarded. other proposed generate valid signatures for its output from input ones...
Network coding is a new technique which appeared in recent years. By employing the inherent broadcast nature of wireless channel, it can achieve higher network throughput networks. Butterfly model depicts basic component unit local area networks (WLANs). In this paper, we propose an algorithm MAC-based specific to - MBNC (MAC-Based Coding). According differences numbers buffered packets for upstream flows node's FIFO output queue, increase opportunity as largely possible by dynamically...
Abstract Quantum computing is an emerging technology which will influence the future. currently limited by scalability of qubits, especially qubits manipulation superconducting computing. In this paper, a method based on AWG (Arbitrary Waveform Generator) proposed. Based quantum system, overall scheme presented, including framework, multichannel coherent arbitrary waveform generation design, state reading and acquisition synchronization design quick feedback transceiver design. The...
Real-world, long-running Internet of Things (IoT) requires intense user-node interaction in the deployment, network operation, and maintenance stages. The rapidly increasing number IoT nodes urges a new user-friendly interface to reduce complexity. This paper presents INSIGHT, an AR-enabled user for allowing users directly grasp perceptual node information from surrounding videos shot by mobile AR devices such as smart glasses phones. INSIGHT incorporates camera, magnetic, IMU sensor data...
Recently, network coding has been applied to reliable multicast in wireless networks improve the transmission efficiency. The expected value of gain, i.e., ratio average number transmissions per packet when using conventional non-coding that coding-based transmission, is an important metric understand benefit from coding. It notable, however, direct calculation gain computationally infeasible. This paper introduces efficient estimation for with much lower computation cost. A formula also...