A5100462265- Music and Audio Processing
- Remote Sensing and Land Use
- Cooperative Communication and Network Coding
- Advanced MIMO Systems Optimization
- Speech and Audio Processing
- Remote-Sensing Image Classification
- Antibiotic Resistance in Bacteria
- Advanced Wireless Network Optimization
- Anomaly Detection Techniques and Applications
- Bacteriophages and microbial interactions
- Industrial Vision Systems and Defect Detection
- Advanced Measurement and Metrology Techniques
- Remote Sensing in Agriculture
- Robotics and Sensor-Based Localization
- Smart Agriculture and AI
- Salmonella and Campylobacter epidemiology
- Animal Vocal Communication and Behavior
- Internet Traffic Analysis and Secure E-voting
- Underwater Vehicles and Communication Systems
- Image and Object Detection Techniques
- Network Security and Intrusion Detection
- Vehicle License Plate Recognition
- Diverse Musicological Studies
- Automated Road and Building Extraction
- Neuroscience and Music Perception
Central South University
2016-2024
Central South University of Forestry and Technology
2016-2024
Shanghai Municipal Center For Disease Control Prevention
2023-2024
Institute of Computing Technology
2024
Chinese Academy of Sciences
2024
Nanjing University of Information Science and Technology
2023
Guangxi University
2022
Shenzhen University
2022
Southwest China Institute of Electronic Technology
2022
Communication University of China
2021
Extracting roads from aerial images is an issue that has attracted much attention. Using semantic segmentation methods to extract often faces the problem of narrow and occluded roads. In this letter, we propose a network called ConDinet++, which improves general codec architecture. encoder part, VGG16 with pretraining parameters utilized for feature extraction. decoder perform fusion mechanism on full-scale map. order improve ability integrate information further increase receptive field,...
For the motion control of industrial robots, end-effector performance is ultimate interest. However, robots are generally only equipped with motor-side encoders. Accurate estimation position and velocity thus difficult due to complex joint dynamics. To overcome this problem, paper presents an optical sensor based on sensitive detector (PSD), referred as PSD camera, for direct sensing. features high precision fast response while being cost-effective, favorable real-time feedback applications....
To solve the complex background problems (e.g. Noise interference, object overlap, and different illumination) that affect classification performance on plant leaf images, this paper proposes an instance segmentation method for images based IFPN SNMS CFFI-Mask R-CNN (ISC-MRCNN) ACPSOSVM-Dual Channels Convolutional Neural Network (APS-DCCNN). obtain foreground of lateral connection structure feature map pyramid in ISC-MRCNN fuses maps depths, so network learns more detailed features. Then,...
At present, many Deep Neural Network (DNN) methods have been widely used for hyperspectral image classification. Promising classification results obtained by utilizing such models. However, due to the complexity and depth of model, increasing number model parameters may lead an overfitting especially when training data are insufficient. As performance mainly depends on sufficient a large network with reasonably optimized hyperparameters, using DNNs requires better hardware conditions time....
There are two major questions regarding Environmental Sound Classification (ESC). What is the best audio recognition framework, and what most robust feature? For investigating above problems, Gated Recurrent Unit (GRU) network was used to analyze effect of single features such as Mel Scale Spectrogram (Mel), Log-Mel (LM), frequency cepstral coefficient (MFCC) well multi-feature about Mel-MFCC, LM-MFCC, Mel-LM-MFCC (T-M) in this paper. The experiment results show that ESC tasks,...
For robots with joint elasticity, discrepancies exist between the motor side and load (e.g., link of robotic joint). Thus, (end-effector) performance can hardly be guaranteed measurements alone. In this paper, a computationally efficient state estimation scheme is proposed for multi-joint robot which equipped encoders low-cost end-effector MEMS sensor such as three-axial accelerometer. An optimization-based inverse differential kinematics algorithm developed to obtain rough estimates. With...
Fifty traditional Tibetan ghee (TTG) varieties were collected and analyzed their systematic characteristic indices, including physicochemical parameters, minerals, fatty acid composition, thermal behavior. Results show that TTG contains a large amount of fat (71.68%-93.3%) small quantity protein (0.51%-1.81%). The peroxide values vary from 0.02 to 1.30 mg/g 0.07 5.93 meq/kg, respectively. content minerals varied with altitude level region significantly (p < 0.05), the regional variations...
In this paper, we investigate the hierarchical clustering for dense small cells and devise non-cooperative game-theoretic scheme, with aim at increasing network throughput minimizing both cross- co-tier interference. By studying distances between requirements of cluster formation, propose a novel scheme densely deployed cells, which consists two phases, i) an absorption mechanism is designed to form into clusters; ii) balance populations, segmentation algorithm proposed clusters that...
Demodulation techniques are of central importance for achieving intelligent receiving. Improvement in demodulation performance enhances the overall a communication system correspondingly. However, conventional demodulators require dedicated hardware platforms leading to high implementation costs and time-consuming development. This work proposes unified architecture end-to-end automatic demodulated modulated signals. The proposed demodulator utilizes residual unit fully convolutional network...
Determining the soil source is crucial for agricultural planning, forensic case analysis, and archaeological site research. LIBS combined with feature selection machine learning algorithms can accurately predict sources.
The plasmid-mediated
Heterogeneous small cell networks (HSCN), as a promising paradigm to increase end-user data rates and improve the overall capacity, is expected be key cellular architecture in 5G wireless networks. However, energy consumed HSCN considerable due massive use of cells. In this paper, we investigate consumption issue which stems from enormous number running base stations (SBSs) deploying HSCN. We first propose power models so characterize active state idle SBSs, respectively. Then two sleep...
Classic vision-based navigation solutions, which are utilized in algorithms such as Simultaneous Localization and Mapping (SLAM), usually fail to work underwater when the water is murky quality of recorded images low. That because most SLAM feature-based techniques often it impossible extract matched features from blurry images. To get more useful features, image processing can be used dehaze before they a navigation/localization algorithm. There many well-developed methods for restoration,...
Device-to-Device (D2D) communications and small cell networks, as promising technologies to improve spectral efficiency system throughput for future cellular have received increasing attentions. In this paper, we model D2D in the two-tier heterogeneous macro-small networks. We propose not only two novel resource sharing strategies users, namely dedicated strategy cross-tier strategy, but theirs corresponding power control well. order minimize interference formulate optimization problem of...
Land vehicle localization and navigation mainly relies on the Global Positioning System (GPS)/Inertial Navigation (INS) integration. In this paper, we propose a unified incremental regression framework that enables with high accuracy in urban environment. Within framework, nonlinear Gauss Process Regression (GPR) approach to perform position prediction during GPS outages. By mapping data into high-dimensional space by kernel function, proposed GPR based is able deal nonlinearity issue denied...
Salmonella is a major cause of food poisoning, and its infection antimicrobial resistance vary regionally due to different sanitary standards the use antimicrobials. The aim this study evaluate microbiological contamination in Shanghai, providing references for supervision over safety clinical treatment. This analysed distribution serotypes, poultry meats serotype Salmonella, correlation between resistance, multiple drug (MDR) 139 strains from meats. results showed that proportion detected...