A5101721113- Rough Sets and Fuzzy Logic
- Advanced Computational Techniques and Applications
- Advanced Neural Network Applications
- Medical Image Segmentation Techniques
- Data Mining Algorithms and Applications
- AI in cancer detection
- IoT and Edge/Fog Computing
- Higher Education and Teaching Methods
- Sustainable Supply Chain Management
- Environmental Impact and Sustainability
- Data Management and Algorithms
- Recycling and Waste Management Techniques
- Advanced Image Processing Techniques
- Image Processing and 3D Reconstruction
- Anomaly Detection Techniques and Applications
- Smart Agriculture and AI
- Multi-Criteria Decision Making
- Technology and Security Systems
- Animal Nutrition and Physiology
- Image Processing Techniques and Applications
- Image and Signal Denoising Methods
- Advanced Text Analysis Techniques
- Imbalanced Data Classification Techniques
- Advanced Image and Video Retrieval Techniques
- Domain Adaptation and Few-Shot Learning
Beijing University of Posts and Telecommunications
2019-2025
Xinjiang Agricultural University
2025
Chengdu University
2015-2024
Central South University
2024
Dongguan University of Technology
2024
Xi'an Jiaotong University
2023-2024
Capital Medical University
2024
Wuhan University
2024
Peking University Third Hospital
2023
State Key Laboratory of Networking and Switching Technology
2023
In real cases, missing values tend to contain meaningful information that should be acquired or analyzed before the incomplete dataset is used for machine learning tasks. this work, two algorithms named jointly fuzzy C-Means and vaguely quantified nearest neighbor (VQNN) imputation (JFCM-VQNNI) fitted VQNN (JFCM-FVQNNI) have been proposed by considering clustering conception sufficient extraction of uncertain information. JFCM-VQNNI JFCM-FVQNNI algorithm, value regarded as a decision...
Abstract Semantic Segmentation has been widely used in a variety of clinical images, which greatly assists medical diagnosis and other work. To address the challenge reduced semantic inference accuracy caused by feature weakening, pioneering network called FTUNet (Feature-enhanced Transformer UNet) was introduced, leveraging classical Encoder-Decoder architecture. Firstly, dual-branch Encoder is proposed based on U-shaped structure. In addition to employing convolution for extraction, Layer...
Utilizing neuroimaging and machine learning (ML) to differentiate schizophrenia (SZ) patients from normal controls (NCs) for detecting abnormal brain regions in has several benefits can provide a reference the clinical diagnosis of schizophrenia. In this study, structural magnetic resonance images (sMRIs) SZ NCs were used discriminative analysis. This study proposed an ML framework based on coarse-to-fine feature selection. The two-sample t -tests extract differences between groups first,...
Preventative effects of Lactobacillus fermentum and Bacillus coagulans against Clostridium perfringens infection in broilers have been well-demonstrated. The present study was conducted to investigate the modulation these two probiotics on intestinal immunity microbiota C. -challenged birds. 336 one-day-old were assigned four groups with six replicates each group. Birds control unchallenged fed a basal diet, birds three challenged dietary supplemented nothing (Cp group), 1 × 10 9 CFU/kg L....
Multi-task semantic communication can serve multiple learning tasks using a shared encoder model. Existing models have overlooked the intricate relationships between features extracted during an encoding process of tasks. This paper presents new graph attention inter-block (GAI) module to encoder/transmitter multi-task system, which enriches for by embedding intermediate outputs in features, compared existing techniques. The key idea is that we interpret feature extraction blocks as nodes...
Transforming recorded videos into concise and accurate textual summaries is a growing challenge in multimodal learning. This paper introduces VISTA, dataset specifically designed for video-to-text summarization scientific domains. VISTA contains 18,599 AI conference presentations paired with their corresponding abstracts. We benchmark the performance of state-of-the-art large models apply plan-based framework to better capture structured nature Both human automated evaluations confirm that...
This paper addresses the dual challenges of food security and sustainable development by examining balance between arable land quality economic development. Coordinating optimizing models is essential for achieving agricultural progress. The North Slope Economic Belt Tianshan Mountain (UANST), a semi-arid agriculturalpastoral transition zone in northwest China, exemplifies coupled human environment system where global sustainability targets confront regional imperatives. Focusing on seven...
Magnetic resonance (MR) images are usually limited by low spatial resolution, which leads to errors in post-processing procedures. Recently, learning-based super-resolution methods, such as sparse coding and convolution neural network, have achieved promising reconstruction results scene images. However, these methods remain insufficient for recovering detailed information from low-resolution MR due the size of training dataset.To investigate different edge responses using kernel sizes, this...
This paper focused on the basic framework of intelligent urban Traffic Management System Based Cloud Computing and Internet Things, proposed architecture Things. The made a deep research information monitoring based internet things, calculation modeling components knowledge matching component. Mass was realized by application cloud computing platform. system fundamentally realizes management traffic purpose dredge traffic.
The objective of the present study was to investigate effects tannic acid (TA) on growth performance, blood parameters, antioxidant capacity, and intestinal health in broilers challenged with aflatoxin B1 (AFB1). A total 480 aged 1 d were randomly allotted into four treatments: 1) CON, control diet; 2) AF, CON + 60 μg/kg AFB1 feed during days 21, 120 22 42; 3) TA1, AF 250 mg/kg TA; 4) TA2, 500 TA. Average daily gain (ADG) average intake (ADFI) increased TA1 42, 42 compared treatments (P <...
ABSTRACT The performance of regional groundwater level (GWL) prediction model hinges on understanding intricate spatiotemporal correlations among monitoring wells. In this study, a graph convolutional network (GCN) with long short-term memory (LSTM) (GCN–LSTM) is introduced for GWL utilizing data from 16 wells located in the northeastern Xiangtan City, China. This designed to account both hybrid temporal dependencies and spatial autocorrelations It consists two parts: part employs GCNs...
The Internet of Things is stepping out its infancy into full maturity, requiring massive data processing and storage. Unfortunately, because the unique characteristics resource constraints, short-range communication, self-organization in IoT, it always resorts to cloud or fog nodes for outsourced computation storage, which has brought about a series novel challenging security privacy threats. For this reason, one critical challenges having numerous IoT devices capacity manage them their...
Existing image segmentation methods are mostly confined to the 2D plane, which only considers information in one direction. As a classic framework with CNN framework, U-Net needs more improvement accuracy. Moreover, 3D CNNs require heavy computational cost. In order balance between accuracy and cost, this paper mainly proposes 2.5D method based on for prediction of cancer's area MRI nasopharyngeal carcinoma. The has sampled patches from images three orthogonal directions, then fed into...
Nasopharyngeal Carcinoma (NPC) is one of the most common malignant tumors in China. However, cancer's region subtle, variability and irregular. In traditional diagnostic way, clinicians' diagnosis relies on manual delineations which are time consuming require rich prior experience. Recently, deep learning architecture U-Net Dual Path Network (DPN) apply well biomedical segmentation nature scene respectively. cannot extract abundance texture information from data DPN utilize shallow layer...