A5103128811- Simulation and Modeling Applications
- Indoor and Outdoor Localization Technologies
- Advanced Computational Techniques and Applications
- Robotics and Sensor-Based Localization
- RFID technology advancements
- Municipal Solid Waste Management
- Underwater Vehicles and Communication Systems
- Energy Harvesting in Wireless Networks
- Geological Modeling and Analysis
- Natural Language Processing Techniques
- Traffic Prediction and Management Techniques
- Smart Materials for Construction
- Transportation Safety and Impact Analysis
- Traffic control and management
- Civil and Geotechnical Engineering Research
- Multi-Criteria Decision Making
- Antenna Design and Analysis
- Topic Modeling
- Infant Nutrition and Health
- Muscle activation and electromyography studies
- Anomaly Detection Techniques and Applications
- Recycling and Waste Management Techniques
- Clinical Nutrition and Gastroenterology
- Acoustic Wave Resonator Technologies
- Asphalt Pavement Performance Evaluation
Third Affiliated Hospital of Zhengzhou University
2023-2025
Tianjin Economic-Technological Development Area
2024
Airport Shunyi District Hospital
2024
Anhui Agricultural University
2023-2024
Sun Yat-sen University
2023-2024
Huazhong University of Science and Technology
2015-2023
Hebei University of Technology
2023
Harbin Welding Institute
2023
Guangdong University of Technology
2023
University of Cambridge
2017-2022
The rapid development of open-source large language models (LLMs) has been truly remarkable. However, the scaling law described in previous literature presents varying conclusions, which casts a dark cloud over LLMs. We delve into study laws and present our distinctive findings that facilitate scale two commonly used configurations, 7B 67B. Guided by laws, we introduce DeepSeek LLM, project dedicated to advancing with long-term perspective. To support pre-training phase, have developed...
Abstract Construction waste is unavoidable in the process of urban development, causing serious environmental pollution. Accurate assessment municipal construction generation requires building identification models using deep learning technology. However, this high-quality public datasets for model training and validation. This study utilizes Google Earth GF-2 images as data source to construct a specific dataset landfills Changping Daxing districts Beijing, China. contains 3,653 samples...
Abstract The use of Santulli enterostomy (SE) for necrotizing enterocolitis (NEC) has been limited to a small number studies involving patients and no control group. Our study aimed compare the clinical safety efficacy with those single- or double-lumen neonatal NEC through retrospective cohort study. One hundred ten who met criteria were divided into an SE group (64 patients) conventional (CE) (46 patients). CE underwent enterostomy. There significant differences in complication rates...
The combination of teleoperating robots and the Internet Things (IoT) could be employed in many areas, including remote nursing semi-mechanical control. However, it is known that subjects can quickly encounter physical mental fatigue, which potentially lower accuracy teleoperation. To address this issue, article presents a closed-loop teleoperation system based on multisensory fusion, visual, haptic feedback within IoT framework. Various sensors for electromyography, inertial measurement...
Background/aims The aim of this study was to develop and evaluate digital ray, based on preoperative postoperative image pairs using style transfer generative adversarial networks (GANs), enhance cataractous fundus images for improved retinopathy detection. Methods For eligible cataract patients, colour photographs (CFP) ultra-wide field (UWF) were captured. Then, both the original CycleGAN a modified (C 2 ycleGAN) framework adopted generation quantitatively compared Frechet Inception...
In recent years, assistive wearables technologies based on Internet of Things (IoT) platforms for amyotrophic lateral sclerosis (ALS) patients trigger broad interests. Nevertheless, the user privacy leakage issue, owing to scene camera installed analyze environmental information, hinders further success use ALS patients. To address this in article, a smart human-environment interactive (HEI) environment, including eye motion detection, radio-frequency identification (RFID), and speech...
Oxygen-enriched waste incineration can improve efficiency and reduce pollutant generation, while greenhouse gases emission characteristics of different wastes are still unclear. In this study, actual municipal solid (MSW) typical components at oxygen concentrations moisture content were investigated to reveal the generation mechanism. The results illustrated that when increased from 21% 30%, CO, CH4 HCN emissions decreased by 95%, 95.5% 96.5%, respectively. NO 75% due promotion on fuel rapid...
In this paper, a phase and RSSI-based method for indoor localization of UHF RFID tags is proposed demonstrated. The exploits the received RSSI profile combined with location moving antenna to locate target tags. measurements are collected by mobile robot carrying reader antennas. By analyzing RSSI, which strength can indicate whether signal stable or not, valid be obtained. After obtaining according along trajectory calculated from stationary point curve. orthogonal distance estimated...
A novel RFID inverse synthetic aperture radar (ISAR)-synthetic (SAR) localization method is proposed and demonstrated which exploits reference tags with known locations to estimate the trajectory of a mobile platform locate target unknown locations. system that integrates robot an integrated reader antenna used obtain phase measurements both when moving along its trajectory. The ISAR algorithm estimates ISAR-SAR loop then adjusts estimated combined determine their location. performance...
Background and objective: IBS-D is a common functional bowel disease with complex etiology without biomarker. The pathological physiological basis of focuses on visceral hypersensitivity. However, its epigenetic mechanism remains elusive. Our study aimed to integrate the relationship between differentially expressed miRNAs, mRNAs proteins in patients order reveal hypersensitivity from transcription protein levels provide molecular for discovering biomarkers IBS-D. Methods: intestinal...
Abstract Image quality variation is a prominent cause of performance degradation for intelligent disease diagnostic models in clinical applications. issues are particularly infantile fundus photography due to poor patient cooperation, which poses high risk misdiagnosis. Here, we developed deep learning-based image assessment and enhancement system (DeepQuality) images improve infant retinopathy screening. DeepQuality can accurately detect various defects concerning integrity, illumination,...
Construction waste is an inevitable byproduct of urban renewal, causing severe pressure on the environment, health, and ecology. Accurately estimating production construction crucial for assessing consumption renewal. However, traditional manual estimation methods rely heavily statistical data historical experience, which lack flexibility in practical applications are time-consuming labor-intensive. In addition, their accuracy timeliness need to be improved urgently. Fortunately, with...
Knee osteoarthritis (Knee OA) is a degenerative disease that often perplexes the elderly and whole society, its timely recognition receives interest worldwide. However, traditional imaging examinations cannot reflect dynamic function nor implement long-term monitoring. To address this issue, article suggests piezoresistive-based gait monitoring method to recognize patients with OA by assessing plantar pressure signals during subjects' walking, which mobile, wearable, low-cost, convenient....
Abstract The Inverse Ocean Modeling (IOM) system constructs and runs weak-constraint, four-dimensional variational data assimilation (W4DVAR) for any dynamical model observing array. dynamics the algorithms may be nonlinear but must functionally smooth. user need only provide algorithms, together with an interpolation scheme that relates numerics to observer’s coordinates. All other model-dependent elements of algorithm (see both Chua Bennett), including adjoint generators Monte Carlo...
The current researches on prediction of crowd stampede accident in large-scale activities are mainly early-warning before the and ways single. To solve this problem, paper proposes a real-time model against risk activity places based Naive Bayes Classifier(NBC) from evaluation index algorithm, verifies rationality validity model.
In recent years, the human-robot interfaces (HRIs) based on surface electromyography (sEMG) have been widely used in lower-limb exoskeleton robots for movement prediction during rehabilitation training patients with hemiplegia. However, accurate and efficient hemiplegia remains a challenge due to complex information individual differences. Traditional methods usually use hand-crafted features, which are computationally cheap but can only extract some shallow heuristic information. Deep...
The indoor localization system is widely used in many scenarios. Here, radio frequency identification (RFID) technology has been approved to be an effective solution. Many RFID-based systems either require high accuracy estimate trajectory such as Synthetic Aperture Radar (SAR)-based methods or are highly depend on the density of reference tags. current study demonstrates first application Graph Neural Network (GNN) for 2D using RFID technology. A graph regression approach was predict tag...
The multi-criteria decision problem is to find the most satisfactory solution among many alternatives, taking into account several criteria that may be conflicting. proposed method focuses on person since importance (weight) given each criterion defined according characteristics of person. We use a special structure Bayesian network (BN) based AHP. graph BN and probabilities associated with nodes are designed translate knowledge experts selection an alternative.
This paper proposes a new antenna remoting system for ultra-high frequency (UHF) radio identification (RFID). The consists of central baseband controller and remote subsystem, which are connected using twisted-pair cable. During the inventory session, only signals transmitted over cable, whilst (RF) up down converted in subsystem. RFID provides -94.5 dBm reader sensitivity, can fulfill long-distance detection tasks with very low performance degradation to 150 m Cat5et Compared conventional...