A5100636124- Adaptive Control of Nonlinear Systems
- Distributed Control Multi-Agent Systems
- Advanced Sensor and Control Systems
- Advanced Memory and Neural Computing
- Ferroelectric and Negative Capacitance Devices
- Software Reliability and Analysis Research
- Terahertz technology and applications
- Advanced Image and Video Retrieval Techniques
- Advanced Measurement and Detection Methods
- Neural Networks Stability and Synchronization
- Magnetic Bearings and Levitation Dynamics
- Risk and Safety Analysis
- Reliability and Maintenance Optimization
- Semiconductor materials and devices
- Radiomics and Machine Learning in Medical Imaging
- Lignin and Wood Chemistry
- Dynamics and Control of Mechanical Systems
- Advanced Decision-Making Techniques
- Advanced Vision and Imaging
- Stability and Control of Uncertain Systems
- Control and Dynamics of Mobile Robots
- Robotics and Sensor-Based Localization
- Advancements in Semiconductor Devices and Circuit Design
- 3D IC and TSV technologies
- Visual Attention and Saliency Detection
Harbin University of Science and Technology
2008-2024
Hebei University
2020-2024
Fujian Electric Power Survey & Design Institute
2023
Hubei University
2022
East China Normal University
2018-2021
Center for High Pressure Science & Technology Advanced Research
2021
Yanshan University
2016-2019
Beihua University
2012-2019
Peking University
2019
Chongqing University of Posts and Telecommunications
2018
BackgroundArtificial intelligence (AI) models in real-world implementation are scarce. Our study aimed to develop a CT angiography (CTA)-based AI model for intracranial aneurysm detection, assess how it helps clinicians improve diagnostic performance, and validate its application clinical implementation.MethodsWe developed deep-learning using 16 546 head neck CTA examination images from 14 517 patients at eight Chinese hospitals. Using an adapted, stepwise evaluation, 120 certified 15...
Abstract In this paper, we propose a saliency detection model for RGB-D images based on the deep features of RGB and depth within Bayesian framework. By analysing 3D in case images, class-conditional mutual information is computed measuring dependence extracted using convolutional neural network; then, posterior probability formulated by applying Bayes’ theorem. assuming that are Gaussian distributions, discriminative mixed-membership naive Bayes (DMNB) used to calculate final map. The...
Face recognition technology is an application of information security, which a kind multi-disciplinary technology, such as comprehensive mathematics, pattern recognition, and biological characteristics. With the development applications, requirements for accuracy anti-counterfeiting face are also increasing. In this paper, K-mean algorithm used to analyze features. Firstly, biometric features extracted, then method cluster Lastly, SVM classify. The results show that can achieve high...
In the millimeter wave (mmW) frequency range, root mean square height of through silicon via (TSV) sidewall roughness is comparable to skin depth, and hence, becomes a critical factor for TSV modeling analysis. this paper, impact on electrical performance, such as loss impedance alteration in mmW examined analyzed. The second-order small analytical perturbation method applied obtain simple closed-form expression power absorption enhancement TSV. study, we propose an model TSV, which...
Scaling down the transistor to gain more computation power will eventually reach unsurmountable physical limitation. Neuromorphic Computing is a novel and promising computing scheme emulating nervous structure data processing methodology of human brain. This paper presents comparison between conventional Von Neumann architecture neuromorphic architecture. We propose 3D combining nanoscale device "memristor" monolithic integration technology, which could reduce system consumption, provide...
Finite-time consensus problem for multiple Euler Lagrange systems under Markovian switching topologies is investigated through an event-triggered control protocol. First, instead of the general stochastic topology, graph entire system governed by some Markov chains to edge set, which can recover traditional in line with practical communication environment. Then, utilising integral sliding-mode strategy, rigorous analysis finite-time performed theory and Lyapunov stability theory. An law...
Recent developments in humidity sensors have heightened the need for reliability. Seeing as many products such experience multiple dependent competing failure processes (MDCFPs) with self-recovery, this paper proposes a new general reliability model. Previous research into MDCFPs has primarily focused on of degradation and random shocks, which are appropriate most products. However, existing models cannot fully characterize significant leading to an underestimation In paper, effect...
Abstract The impact of hypochlorous acid (HOCl) treatment on a fully bleached eucalyptus kraft pulp was studied with respect to the final brightness as well reversion under humid ageing conditions. carbonyl contents in pulps were determined by CCOA method where carbazole-9-carboxylic [2-(2-aminooxethoxy)-ethoxy] amide (CCOA) applied carbonyl-selective fluorescence label. molecular weight distribution cellulose and hemicelluloses analyzed well. amount groups observed increase significantly...
3D integration technology offers a near term strategy for bypassing Moore's Law. Applying to neuromorphic computing (NC) could provide low power consumption, high-connectivity, and massively parallel processed system that can accommodate high demand computational tasks. This paper proposes novel analog spiking nanoscale NC system, wherein both neurons synapses are stacked three-dimensionally, with monolithic inter-tier via (MIV) technology, vertical resistive random-access memory (V-RRAM)...
Three-dimensional (3D) integrated circuits (ICs) offer a promising near-term solution for pushing beyond Moore's Law because of their compatibility with current technology while providing high system speed, density, massively parallel processing, low power consumption, and small footprint. In this paper, novel 3D neuromorphic IC architecture combining monolithic integration vertical resistive random-access memory (V-RRAM) is proposed. Furthermore, concise equivalent circuit model the...
Recent advances in ultrafast optical laser technology have improved generation and detection of energy within the terahertz (THz) portion electromagnetic (EM) spectrum. One promising application THz spectroscopy is explosive materials chemical or biological agents. This has been motivated by initial measurements that indicate explosives may unique spectral characteristics region thus providing a discernible fingerprint. However, since wavelengths are 10's to 100's microns scale, rough...
Proposed is an optimised Linux swap system, called SGBI, where the kernel manages flash memory-based area directly without FTL. This approach not only introduces efficient swap-aware garbage collection policy to reduce its cleaning cost and improve degree of wear-levelling, but also adopts swap-in read-ahead scheme delay execution collection. Swap I/O traces from have been collected trace-driven simulations performed, which show that proposed greatly outperforms existing systems.
Bayesian Network has an advantage in dealing with uncertainty. But It is difficult to construct a scientific and rational model practice application. In order solve this problem, novel method for constructing by integrating Failure Mode Effect Analysis (FMEA) Fault Tree (FTA) was proposed. Firstly, the structure matrix representations of FMEA, FTA were shown integration algorithm explained. Then, approach obtaining information on node, parameter from FMEA based put forward. Finally, verify...
Shape memory alloys (SMAs) in the form of rings or tubes with two-way shape effect (TWSME) have been applied to high-speed precision micro-mechanical systems. This study investigates development NiTi alloy TWSME and analyzes effects annealing thermo-mechanical treatments. The results show there is a maximum respect temperature, loading deformation, training cycle, respectively. SMA stabilizes earlier than structure dimension during training. change phase transition temperatures, which has...
In this paper, we proposed to use 3D integration technology create a neuromorphic hardware system that is compatible with current technology, provides high speed, density, massively parallel processing, low power consumption, and small footprint. The Through Silicon Vias (TSVs) used in the structure provide density energy efficient links for transferring information through multiple neuron layers. This work details how benefited from redundant TSV substantial design-area reduction. We...
Abstract This paper presents a distributed consensus algorithm that employs event‐triggered communication for multiple underactuated systems under Markovian switching topologies. Instead of the general stochastic topology, graph entire system is governed by set Markov chains to edges, which can recover topologies in line with practical network. By utilizing integral sliding mode control strategy, rigorous analysis asymptotic convergence results has been performed through theory and Lyapunov...
With the improvements in sensor technologies and image processing algorithms, computer vision has become a major tool for robots to recognize gauge their surroundings. For instance, Kinect can be used as an excellent depth camera indoor navigation. However, there exist situations that need recognition spatial interpretation of environment using limited hardware resources. The is not suitable outdoor use while LIDAR too large expensive installed on autonomous miniature surveillance drone....
In this paper, we propose a saliency detection model for RGB-D images based on the contrasting features of color and depth within Bayesian framework. The feature map is extracted superpixel contrast computation with spatial priors. We by approximating density depth-based using Gaussian distribution. Similar to computation, computed distribution multi-scale contrasts in superpixels exploiting low-level cues. By assuming that color- are conditionally independent, given classes, discriminative...
In the present work, a frequency-agile antenna is obtained by integrating simple planar inverted F (PIFA) with programmable impedance tuner. The tuner, software-defined match control (SDMC) circuit, uses microstrip loaded line structure three stubs that are each terminated varactor diode. spectral response of input varying voltage bias and thus capacitance diodes. tuner only controls between PIFA 50-ohm design internal RF circuits.
The translucency of terahertz wave through many polymer composite materials, particularly foams, makes it a good complement to another nondestructive testing (NDT) methods. Terahertz time-domain spectroscopy (THz-TDS) technology can obtain amplitude and phase information simultaneously. It is powerful tool for foam materials embody the depth defects while acquiring defect location, size, nature. In process NDT, refractive index one important factors that affect detection effect. previous...