A5100443824- Advanced Malware Detection Techniques
- Music and Audio Processing
- Speech and Audio Processing
- Network Security and Intrusion Detection
- Time Series Analysis and Forecasting
- Music Technology and Sound Studies
- Aerodynamics and Acoustics in Jet Flows
- Anomaly Detection Techniques and Applications
- Software Testing and Debugging Techniques
- Face recognition and analysis
- Advanced Manufacturing and Logistics Optimization
- Biometric Identification and Security
- Advanced Vision and Imaging
- Neural Networks and Applications
- Software Engineering Research
- Advanced Algorithms and Applications
- Artificial Intelligence in Games
- Indoor and Outdoor Localization Technologies
- Security and Verification in Computing
- Video Coding and Compression Technologies
- Diverse Musicological Studies
- Research studies in Vietnam
- Advanced Image Processing Techniques
- Blind Source Separation Techniques
- Advanced Image and Video Retrieval Techniques
China Tobacco
2024
State Key Laboratory of Vehicle NVH and Safety Technology
2024
Nanchang Hangkong University
2024
Harbin Engineering University
2021-2022
China Southern Power Grid (China)
2020-2021
Wuhan Engineering Science & Technology Institute
2020
NetEase (China)
2020
Shandong University
2019
National University of Defense Technology
2012
Yangtze University
2008
This paper presents a novel multitask learning framework for palmprint biometrics, which optimizes classification and hashing branches jointly. The branch within our facilitates the concurrent execution of three distinct tasks: identity recognition soft encompassing gender chirality. On other hand, enables generation hash codes, optimizing minimal storage as templates efficient matching. derives complementary information from these tasks by amalgamating knowledge acquired branch. approach...
Rongsheng Zhang, Xiaoxi Mao, Le Li, Lin Jiang, Chen, Zhiwei Hu, Yadong Xi, Changjie Fan, Minlie Huang. Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing: System Demonstrations. 2020.
Transformer-based methods have recently demonstrated their potential in time series forecasting problems. However, the mainstream approach, primarily utilizing attention to model inter-step correlation domain, is constrained by two significant issues that lead ineffective and inefficient multivariate forecasting. The first key representations domain are scattered sparse, resulting parameter bloat increased difficulty capturing dependencies. second treating step points as uniformly embedded...
The article analyzes the advantage and disadvantage of part lacking fidelity searching algorithm, based on early ending technology thought. Given partial distortion search (PDS) it brings up one kind improved method which is similar to it. algorithm also calculates sum absolute difference (SAD) once within a row add block pixel SAD, but two disadvantages has been improved. results related experiment show that termination fine granularity adaptive threshold points in levels introduction...
Virtual machine monitor (VMM)-based anti-malware systems have recently become a popular research topic in finding ways of overcoming the fundamental limitations traditional host-based systems, which are likely to be deceived and attacked by malicious codes. This paper analyzes existing VMM-based models malware detection. "Out-of-the-box" detection, active defense model, or In-VM same defects: (1) on top VMM, two virtual machines used, one user (Guest OS) other as (Host OS), (2) users cannot...
In recent years, malicious programs seriously threaten the security of information system. Because its particularity, complexity and vulnerability, power system is difficult to detect kill by traditional anti-virus software. To solve above problems, this paper proposes a behavior detection method based on deep learning, which can identify attack types according activities software behaviors. paper, hybrid learning structure convolutional neural network (CNN) long-and-short term memory (LSTM)...
Few-shot audio classification is an emerging topic that attracts more and attention from the research community. Most existing work ignores specificity of form spectrogram focuses largely on embedding space borrowed image tasks, while in this work, we aim to take advantage special format propose a new method by hallucinating high-frequency low-frequency parts as structured concepts. Extensive experiments ESC50 our curated balanced Kaggle18 dataset show proposed outperforms baseline notable...
Along with the rapid development of information technology and industrial IoT, a large amount time series data are continuously produced in widespread fields, including medicine, finance, astronomy so on. Accordingly, mining has widely followed by researchers classification (TSC) also been attracting great interest over past decade. Recent empirical evidence suggested that combination nearest neighbor classifiers (NNC) Dynamic Time Warping (DTW) is more efficiently than traditional methods....
Abstract The emergence of edge computing technology brings new development direction to the construction power grid industry. At present, for China is still in exploratory stage, combination and may face security risks. This paper analysed risks that occur under circumstances identification system, communication technology, equipment protection resource constraints, put forward relevant risk analysis model defence, aiming at solving problems promoting better grid.
Few-shot audio classification is an emerging topic that attracts more and attention from the research community. Most existing work ignores specificity of form spectrogram focuses largely on embedding space borrowed image tasks, while in this work, we aim to take advantage special format propose a new method by hallucinating high-frequency low-frequency parts as structured concepts. Extensive experiments ESC-50 our curated balanced Kaggle18 dataset show proposed outperforms baseline notable...
Fast and accurate acoustic source localization methods have essential application value in the field of aircraft. Compared with traditional model-based methods, technology based on deep learning shows a good prospect. However, uninterpretability limits further development this technology. This paper proposes network fast iterative shrinkage threshold algorithm unfolding (FISTA-Net), which combines advantages methods. In FISTA-Net, steps are mapped into network, model parameters can be...
Tympanometry has become a standard evaluation procedure for children and adults in the audiological clinic. Because of acoustical complexity both middle ear system testing device, there is need to study tympanometry not only from an experimental but also theoretical perspective achieve full understanding underlying mechanism. In present study, model been developed simulation different type tympanograms. model, acoustic characteristics canal are described by set equations those Zwislocki...