A5100668302- Text and Document Classification Technologies
- Advanced Image Fusion Techniques
- Advanced Image Processing Techniques
- Neural Networks and Applications
- Machine Learning in Bioinformatics
- Advanced Neural Network Applications
- Adversarial Robustness in Machine Learning
- Machine Learning and ELM
- Image and Signal Denoising Methods
- RNA and protein synthesis mechanisms
- Advanced Clustering Algorithms Research
- Advanced Sensor and Control Systems
- Sentiment Analysis and Opinion Mining
- Image Retrieval and Classification Techniques
- Brain Tumor Detection and Classification
- Advanced Graph Neural Networks
- Machine Learning in Materials Science
- Advanced Image and Video Retrieval Techniques
- Underwater Vehicles and Communication Systems
- Face and Expression Recognition
- Domain Adaptation and Few-Shot Learning
- Web Data Mining and Analysis
- Topic Modeling
- Rough Sets and Fuzzy Logic
- Water Quality Monitoring Technologies
Prairie View A&M University
2019-2024
Zhejiang University of Technology
2023
East China Normal University
2023
China Institute of Water Resources and Hydropower Research
2011
Recently, human being’s curiosity has been expanded from the land to sky and sea. Besides sending people explore ocean outer space, robots are designed for some tasks dangerous living creatures. Take exploration an example. There many projects or competitions on design of Autonomous Underwater Vehicle (AUV) which attracted interests. Authors this article have learned necessity platform upgrade a previous AUV project, would like share experience one task extension in area fish detection....
Online Knowledge Distillation (OKD) is designed to alleviate the dilemma that high-capacity pre-trained teacher model not available. However, existing methods mostly focus on improving ensemble prediction accuracy from multiple students (a.k.a. branches), which often overlook homogenization problem makes student saturate quickly and hurts performance. We assume intrinsic bottleneck of comes identical branch architecture coarse strategy. propose a novel Adaptive Hierarchy-Branch Fusion...
Recently, the deep learning based methods, especially ones on convolutional neural network (CNN), achieved remarkable progresses in sentiment analysis. However, CNN methods do not take latent topic text into consideration. In this paper, we propose a Diversified Restrict Boltzmann Machine (RBM) method to model sequence level topics sentences for The basic idea is use mapping feature space, and then utilize RBM text. obtained features are embedding improve performance of evaluations COAE 2014...
One of the main factors that affect supervised learning performance neural network (NN) is training data set. However, for real world problems, it not always easy to obtain high quality sets. In this paper, a novel selection method based on shadowed sets proposed can select an informative and representative subset ameliorate NN. The goal work improve generalization ability diminish misclassification errors classifier This paper firstly introduces central idea Then algorithm described in...
In this paper, we propose a machine learning based recommendation engine that can assist embedded programmers to quickly search and query related code segments instructions. the design, build database each segment is auto-classified multi-labeled with chip model, application scenario, function module, register name. The similarity of level calculated by Hamming Distance, Cosine similarity, Euclidean distance approaches separately. Based on different similarities level, come up dynamic model...
With the development of Internet Things (IoT) technology, embedded based electronic devices have penetrated every corner our daily lives.As brain IoT devices, micro controller unit (MCU) plays an irreplaceable role.The functions MCUs are becoming more and powerful complicated, which brings huge challenges to programmers.Embedded code, is highly related hardware resources, differs from other popular programming code.The configuration may be a big challenge programmers, who only good at...
Identifying DNA N6-methyladenine (6mA) sites is significantly important to understanding the function of DNA. Many deep learning-based methods have been developed improve performance 6mA site prediction. In this study, further prediction, we propose a new meta method, called Co6mA, integrate bidirectional long short-term memory (BiLSTM), convolutional neural networks (CNNs), and self-attention mechanisms (SAM) via assembling two different models. The first model in study CBi6mA, which...
Satellite images are essential for providing geoinformation in Earth Science. Limited imaging devices make high-resolution hard to obtain, bringing difficulties precise predictions many applications. In this paper, a novel unpaired TransCycleGAN network is proposed super-resolve remote sensing using pseudo-supervision. Specifically, based on CycleGAN with an integrated effective degradation-removal Transformer module. Benefiting from pseudo-supervision, clean and naturally degraded...
While graph neural networks have achieved state-of-the-art performances in many real-world tasks including classification and node classification, recent works demonstrated they are also extremely vulnerable to adversarial attacks. Most previous focused on attacking under impractical white-box scenarios. In this work, we will propose a non-targeted Hard Label Black Box Node Injection Attack Graph Neural Networks, which the best of our knowledge, is first its kind. Under setting, more real...