A5001937116- Music and Audio Processing
- Speech and Audio Processing
- Speech Recognition and Synthesis
- Machine Learning and Algorithms
- EEG and Brain-Computer Interfaces
- Face and Expression Recognition
- Text and Document Classification Technologies
- Blind Source Separation Techniques
- Remote Sensing and Land Use
- Advanced Adaptive Filtering Techniques
- Brain Tumor Detection and Classification
- Video Analysis and Summarization
- Advanced Algorithms and Applications
- Mineral Processing and Grinding
- Anomaly Detection Techniques and Applications
- Sentiment Analysis and Opinion Mining
- Ferroelectric and Piezoelectric Materials
- Image Processing Techniques and Applications
- Indoor and Outdoor Localization Technologies
- Geotechnical Engineering and Soil Mechanics
- Music Technology and Sound Studies
- Evaluation and Optimization Models
- Advanced Text Analysis Techniques
- Algorithms and Data Compression
- Geomechanics and Mining Engineering
Shandong Normal University
2015-2024
Changchun University of Chinese Medicine
2024
Guizhou University
2024
Southwest Forestry University
2019
Beijing Jiaotong University
2012-2014
United States Department of Transportation
2014
Jangan University
2012-2014
Shandong University
2005-2006
Automatic epilepsy detection is of great significance for the diagnosis and treatment patients. Most methods are based on patient-specific models have achieved good results. However, in practice, new patients do not their own previous EEG data therefore cannot be initially diagnosed. If other can used to achieve cross-patient detection, experiments combined at same time, this method will more widely used. In work, an classification model a self-organizing fuzzy logic (SOF) classifier...
This paper examines the development of asset pricing models following chronological order. Beginning with traditional methods, this introduces historical background Capital Asset Pricing Model (CAPM) as well strengths and weaknesses application model in real world. To enhance predictiveness CAPM improve over-simplified model, researchers have come up FAMA-French multiple factors. These incorporate more factors when predicting returns assets, but they are still based on linear relationships....
The automatic identification of epileptic electroencephalogram (EEG) signals can give assistance to doctors in diagnosis epilepsy, and provide the higher security quality life for people with epilepsy. Feature extraction EEG determines performance whole recognition system. In this paper, a novel method using local binary pattern (LBP) based on wavelet transform (WT) is proposed characterize behavior activities. First, WT employed time-frequency decomposition signals. After that, "uniform"...
The overhead caused by broadcast schemes cannot be neglected in mobile ad hoc networks. It is known that the probabilistic scheme an effective approach to reduce overhead, which nodes forward packets with a probability redundant retransmissions. In hostile environments, malicious degrade network efficiency, and they should prevented from route discovery data transmission. this case, routing forwarded are redundant. This paper presents Trust-based Probabilistic Broadcast (TPB) focusing on...
Inflammatory cytokines have been linked to digestive system cancers, yet their exact causal connection remains uncertain. Consequently, we conducted a Mendelian randomization (MR) analysis gauge how inflammatory are the risk of five prevalent cancers (DSCs).
Transfer efficiency in integrated transportation terminal is greatly important for both passengers and operational companies. In this paper, we proposed various criteria a hierarchy index system to evaluate the performance of transfer condition inside Beijing South Railway Station. To make assessment more scientific, assign weightings each them by weighting method. Then use an evaluation method, Multi-level Grey Evaluation, calculate indexes different modes station further compare ranking...
In recent years, Transformer has shown great performance in speech enhancement by applying multi-head self-attention to capture long-term dependencies effectively. However, the computation of is quadratic with input spectrograms, which makes it computationally expensive for practical use. this paper, we propose a low complexity hierarchical frame-level Swin network (FLSTN) enhancement. FLSTN takes several consecutive frames as local window and restricts within it, reducing linear spectrogram...
Abstract Subspace-based methods have been effectively used to estimate enhanced speech from noisy samples. In the traditional subspace approaches, a critical step is splitting of two invariant subspaces associated with signal and noise via decomposition, which often performed by singular-value decomposition or eigenvalue decomposition. However, these algorithms are highly sensitive presence large corruptions, resulting in amount residual within low signal-to-noise ratio (SNR) situations....