- Robotics and Sensor-Based Localization
- Advanced Algorithms and Applications
- Target Tracking and Data Fusion in Sensor Networks
- Blind Source Separation Techniques
- Robotic Path Planning Algorithms
- Metaheuristic Optimization Algorithms Research
- Advanced Adaptive Filtering Techniques
- Control Systems and Identification
- Energy Efficient Wireless Sensor Networks
- Image and Signal Denoising Methods
- Medical Image Segmentation Techniques
- Image and Object Detection Techniques
- Advanced Measurement and Detection Methods
- Advanced Sensor and Control Systems
- Artificial Immune Systems Applications
- Advanced Image Fusion Techniques
- Advanced X-ray and CT Imaging
- Underwater Vehicles and Communication Systems
- 3D Surveying and Cultural Heritage
- Distributed Sensor Networks and Detection Algorithms
- Vehicle License Plate Recognition
- Industrial Technology and Control Systems
- Remote Sensing and Land Use
- Mobile Agent-Based Network Management
- Medical Imaging Techniques and Applications
Suizhou Central Hospital
2024
Southwest Jiaotong University
2024
First Affiliated Hospital of Chengdu Medical College
2024
Huawei Technologies (China)
2024
Zhejiang University
2023
Chinese University of Hong Kong
2018-2022
South China Agricultural University
2021
Guangzhou Panyu Polytechnic
2021
Shanghai University
2009-2019
China Jiliang University
2018
Forest fires are serious disasters that affect countries all over the world. With progress of image processing, numerous image-based surveillance systems for have been installed in forests. The rapid and accurate detection grading fire smoke can provide useful information, which helps humans to quickly control reduce forest losses. Currently, convolutional neural networks (CNN) yielded excellent performance recognition. Previous studies mostly paid attention CNN-based classification...
A model of multiwavelet-based neural networks is proposed. Its universal and L/sup 2/ approximation properties, together with its consistency are proved, the convergence rates associated these properties estimated. The structure this network similar to that wavelet network, except orthonormal scaling functions replaced by multiscaling functions. theoretical analyses show multiwavelet converges more rapidly than especially for smooth To make a comparison between both networks, experiments...
Registering two point sets (PSs) is an essential problem in medical robotics and computer-assisted surgery (CAS). As one typical example, orthopedic (CAOS), the preoperative scan has to be aligned with intraoperative accurately. In this article, we first formally formulate generalized PS registration a probabilistic manner. Especially, not only positional but also orientational information incorporated into registration. Notably, error assumed obey multivariate Gaussian distribution...
Abstract With the expansion of freight aerial cableways in mountainous regions, safety monitoring becomes crucial. Traditional methods are inefficient and lack intelligent warning systems. This paper proposes SafeRopeSystem, a deep learning-based system. It utilizes an edge-cloud architecture with SafeRopeNet, lightweight target detection network enhanced by Adaptive Feature Enhancement Module Weather-Robust Pyramid Network for resilience adverse weather. The system includes SafeHead...
We have cloned the gene cluster encoding three subunits of membrane-bound gluconate dehydrogenase (GADH) from Erwinia cypripedii ATCC 29267 in Escherichia coli by performing a direct-expression assay. The positive clone converted D-gluconate to 2-keto-D-gluconate (2KDG) culture medium. Nucleotide sequence analysis GADH revealed that fragment contained complete structural genes for 68-kDa subunit, 47-kDa cytochrome c and 24-kDa subunit unknown function were clustered with same transcriptional...
To achieve scenario intelligence, humans must transfer knowledge to robots by developing goal-oriented algorithms, which are sometimes insensitive dynamically changing environments. While deep reinforcement learning achieves significant success recently, the long and difficult training process limits its application. In this paper, we propose a hybrid structure named Option-Interruption in human is embedded into hierarchical framework. Our architecture has two key components: options,...
The mutation at epidermal growth factor receptor (EGFR) is a clinical predictor of EGFR tyrosine kinase inhibitors (TKI) in patients with non-small cell lung cancer (NSCLC). serum carcinoembryonic antigen (CEA) level was regarded as predictive for the EGFR-TKI efficacy. Are there any other markers? This study analysed correlation between treatment effect and multiple tumor markers only adenocarcinoma to find efficacy.Clinical features, survival time, marker levels before were analysed,...
In image-guided minimally invasive surgery, the clinician relies on image guidance to observe, plan and navigate. order show invisible vessels or planning annotations in live X-ray images, update information for planning, correspondence establishment between 2D images pre-operatively acquired 3D CT is a fundamental step. Accurate alignment needed can be provided by 2D-3D registration bringing pre-operative intra-operative into same coordinate. this work, we propose novel method register...
For airports with high passenger and luggage flows, a large number of staff members have to be hired deploy the scattered trolleys. To release humans from repetitive laborious job, we develop an autonomous trolley deployment robot detect, transport collect idle trolleys recycling points. This paper will firstly illustrate entire collection pipeline system then address key challenge: partial full point set registration. With perception framework, can detect acquire pose on ground, capture...
In this paper, a novel definition of opposite path is proposed. Its core feature that the node sequence candidate paths and distances between adjacent nodes in tour are considered simultaneously. sense, its corresponding have same (or similar, at least) distance from optimal current population. Based on an accepted framework for employing opposition-based learning, Oppositional Biogeography-Based Optimization using Current Optimum, called COOBBO algorithm, introduced to solve combinatorial...
The pth-order inverse method is one of important approaches to Volterra equalization. However, when a equalizer instead an exact connected in cascade before (after) nonlinear system, the existence filter equalization and approximation output error bound resulting system have yet be reported. In this paper, concept local l/sup 2/ stability for introduced, algorithmic formulae via multidimension z-transform are presented. signal investigated as well. It shown that tends zero p infinity finite...
Mobile agent (MA)-based network management and policy-based are new methods to manage the network. They both have their strong points weaknesses. This paper integrated NM into MA-based NM, built a prototype test its performance. We found that with this prototype, we could build an automatic system good security mobility. It would ease of increasing complex
In this study, a systematic solution for mobile manipulation with the application of autonomous trolley collection in airport is presented. The developed system features novel integrated hardware and software system, incorporating specially designed manipulator capable tightly catching trolley. Particularly, we innovatively present supervisory control guided coarse-to-fine visual strategy large-scale environments. At coarsest level, an image-based servoing method utilized to approach from...
Non-rigid point set registration (PSR) is an outstanding and fundamental problem in fields of robotics, computer vision, medical image analysis imageguided surgery. The aim a non-rigid to align together two sets that have been deformed. We derived presented novel algorithm non-rigidly registers together. assumption isotropic localization error shared the previous algorithms. In this paper, position generalized anisotropic cases, which means distribution not same different spatial directions....
Human activities are hugely restricted by COVID-19, recently. Robots that can conduct inter-floor navigation attract much public attention since they substitute human workers to the service work. However, current robots either depend on assistance or elevator retrofitting, and fully autonomous is still not available. As very first step of navigation, button segmentation recognition hold an important position. Therefore, we release large-scale publicly available panel dataset in this work,...
This paper presents a novel, robust, and accurate three-dimensional (3D) rigid point set registration (PSR) method, which is achieved by generalizing the state-of-the-art (SOTA) Bayesian coherent drift (BCPD) theory to scenario that high-dimensional sets(PSs) are aligned anisotropic positional noise considered. Our contributions in this three folds. First, problem of rigidly aligning two general sets (PSs) with normal vectors incorporated into variational inference framework, solved BCPD...
The equalization of nonlinear systems is an important application Volterra filters. pth-order inverse one the approaches equalization, but up to present it has only been applied single-input multiple-output (SISO) case. convergence filter and error bound using approximation equalizer instead have not yet reported. In this paper, we generalize method multi-input multi-output (MIMO) case, discuss equalizer. concept local I/sub 2/ stability for MIMO first introduced, then algorithmic formula...