A5008289463- Face and Expression Recognition
- Machine Learning and ELM
- Advanced Steganography and Watermarking Techniques
- GNSS positioning and interference
- Domain Adaptation and Few-Shot Learning
- Digital Media Forensic Detection
- Sparse and Compressive Sensing Techniques
- Geophysics and Gravity Measurements
- Text and Document Classification Technologies
- Advanced Algorithms and Applications
- Inertial Sensor and Navigation
- Tribology and Lubrication Engineering
- Remote-Sensing Image Classification
- Tribology and Wear Analysis
- Anomaly Detection Techniques and Applications
- Water Quality Monitoring Technologies
- Neural Networks and Applications
- Internet Traffic Analysis and Secure E-voting
- Spectroscopy and Chemometric Analyses
- Video Surveillance and Tracking Methods
- Imbalanced Data Classification Techniques
- Advanced Statistical Methods and Models
- Remote Sensing and Land Use
- Matrix Theory and Algorithms
- Multimodal Machine Learning Applications
China Jiliang University
2025
China Agricultural University
2015-2024
Ministry of Agriculture and Rural Affairs
2022-2024
National Engineering Research Center for Information Technology in Agriculture
2022-2024
Dongfang Electric Corporation (China)
2016-2022
China Medical University
2022
First Hospital of China Medical University
2022
Hunan University
2016
Ministry of Transport
2015
Shanghai Municipal Center For Disease Control Prevention
2015
The prevailing text steganalysis methods detect steganographic communication by extracting hand-crafted features and classifying them using SVM. However, these are designed based on the statistical changes caused steganography, thus they difficult to adapt different kinds of embedding algorithms detection performance is heavily dependent size. In this letter, we propose a novel model convolutional neural network, which able capture complex dependencies learn feature representations...
Text has become one of the most extensively used digital media in Internet, which provides steganography an effective carrier to realize confidential message hiding. Nowadays, generation-based linguistic made a significant breakthrough due progress deep learning. However, previous methods based on recurrent neural network have two deviations including exposure bias and embedding deviation, seriously destroys security steganography. In this article, we propose novel steganographic model...
Abstract We investigate the 3‐D tidal displacement field on Earth's surface recorded globally by 456 continuous global positioning system (GPS) stations of IGS spanning 1996–2011, for eight principal diurnal and semidiurnal constituents. In‐phase quadrature amplitudes residual displacements, after removal an a priori body tide model, are estimated using precise point (PPP) technique daily GPS data; resultant estimates combined to derive final each at station. The results compared with...
With the emergence of generation-based steganography, traditional text steganalysis methods show unsatisfactory detection performance as manually extracted features are simple and non-universal. The recently proposed deep learning-based can obtain great accuracy by extracting high-level features. In this letter, a hybrid method (R-BILSTM-C) is through combining advantages Bidirectional Long Short Term Memory Recurrent Neural Network (Bi-LSTM) Convolutional (CNN). efficiently capture both...
Laser frequency combs (LFCs) are an important component of Doppler radial velocity (RV) spectroscopy that pushes fractional precision to the $10^{-10}$ level, as required identify and characterize Earth-like exoplanets. However, large intensity variations across LFC spectrum arise in nonlinear broadening limit range comb lines can be used for optimal wavelength calibration with sufficient signal-to-noise ratio. Furthermore, temporal spectral-intensity fluctuations LFC, coupled flux-dependent...
Passive daytime radiative cooling (PDRC) presents an effective strategy for mitigating the global greenhouse effect.
Abstract The increasing variety of distributed generation (DG) types introduces significant power fluctuations and uncertainties when integrated into the distribution network. This imposes challenges in ensuring stability safety system. Flexible configuring network’s topology can reduce active losses improve node voltage levels. However, traditional heuristic algorithms often fall short addressing reconfiguration networks with multiple DG types. In this paper, an enhanced Quantum Particle...
Abstract Multi-class classification is an important and on-going research subject in machine learning. Recently, the ν-K-SVCR method was proposed by authors for multi-class classification. As many optimization problems have to be solved classification, it extremely develop algorithm that can solve those efficiently. In this article, problem reformulated as affine box constrained variational inequality with a positive semi-definite matrix, regularized version of nonsmooth Newton uses D-gap...
Feature-based domain adaptation methods project samples from different domains into the same feature space and try to align distribution of two learn an effective transferable model. The vital problem is how find a proper way reduce shift improve discriminability features. To address above issues, we propose unified Probability-based Graph embedding Cross-domain class Discriminative learning framework for unsupervised (PGCD). Specifically, novel graph structures be discriminative transfer...