A5100623585- Infrastructure Maintenance and Monitoring
- Advanced Vision and Imaging
- Robotics and Sensor-Based Localization
- Image Processing Techniques and Applications
- Advanced Measurement and Detection Methods
- Medical Imaging and Analysis
- Industrial Vision Systems and Defect Detection
- Asphalt Pavement Performance Evaluation
- Indoor and Outdoor Localization Technologies
- Image and Object Detection Techniques
- Astronomical Observations and Instrumentation
- Concrete Corrosion and Durability
- Optical Imaging and Spectroscopy Techniques
- Supramolecular Self-Assembly in Materials
- Non-Invasive Vital Sign Monitoring
- Neural Networks and Applications
- Protein Hydrolysis and Bioactive Peptides
- Nanomaterials and Printing Technologies
- Gait Recognition and Analysis
- Advanced MRI Techniques and Applications
- Adaptive Control of Nonlinear Systems
- Surgical Simulation and Training
- Radiomics and Machine Learning in Medical Imaging
- Electromagnetic Simulation and Numerical Methods
- Plant-Microbe Interactions and Immunity
Ministry of Water Resources of the People's Republic of China
2024-2025
Ministry of Natural Resources
2024-2025
Tsinghua University
2012-2024
Hebei University of Technology
2007-2024
Centre National de la Recherche Scientifique
2012-2024
Chimie ParisTech
2024
Université Paris Sciences et Lettres
2024
Nanjing Forestry University
2024
Zhejiang Ocean University
2024
Institut de Recherche de Chimie Paris
2024
Automatic detection of pavement cracks is an important task in transportation maintenance for driving safety assurance. However, it remains a challenging due to the intensity inhomogeneity and complexity background, e.g., low contrast with surrounding possible shadows similar intensity. Inspired by recent success on applying deep learning computer vision medical problems, deep-learning based method crack proposed this paper. A supervised convolutional neural network trained classify each...
Pavement crack detection is a critical task for insuring road safety. Manual extremely time-consuming. Therefore, an automatic method required to boost this progress. However, it remains challenging due the intensity inhomogeneity of cracks and complexity background, e.g., low contrast with surrounding pavements possible shadows similar intensity. Inspired by recent advances deep learning in computer vision, we propose novel network architecture, named feature pyramid hierarchical boosting...
Spatial population distribution data is the discretization of demographic into spatial grids, which has vital reference significance for disaster emergency response, assessment, rescue resource allocation, and post-disaster reconstruction. The random forest (RF) model, as a prominent method modeling population, been studied by many scholars, both domestically abroad. Specifically, research focused on aspects such multi-source fusion, feature selection, accuracy evaluation within process....
The ongoing COVID-19 pandemic, caused by the highly contagious SARS-CoV-2 virus, has overwhelmed healthcare systems worldwide, putting medical professionals at a high risk of getting infected themselves due to global shortage personal protective equipment. This in-turn led understaffed hospitals unable handle new patient influx. To help alleviate these problems, we design and develop contactless positioning system that can enable scanning patients in completely remote fashion. Our key...
Highway crack segmentation is a critical task for highway infrastructure monitoring and maintenance. While imagery from unmanned aerial vehicles (UAVs) applied to the of segmentation, it has great prospects in terms speed range. However, difficult accurately identify road cracks UAV remote sensing images, because are very narrow small, often containing only few pixels. To improve this study proposed an improved identification technique based on U-Net architecture enhanced with convolutional...
Urban nighttime lighting extends human activity hours and enhances safety but also wastes energy causes light pollution. Influenced by building obstructions surface reflections, emissions exhibit significant anisotropy. Remote sensing can be used to observe from high altitudes, ground anisotropy introduces angle-related errors. This study constructed a 3D urban model using virtual simulations conducted multi-angle observations investigate its influencing factors. The results show that the...
Abstract One‐pot processes have emerged as a powerful strategy in (macro)molecular synthesis: integrating multicatalytic process maximizes efficiency, reduces waste, improves profitability, and provides versatile tools for designing of more sustainable without compromising selectivity activity. This review article critical overview the application one‐pot transformations polymerization, with goal synthesizing polymers tailored structures functions wide range applications. Recent advances...
This paper describes the two-dimensional (2D) crystallization of side-chain giant molecules (SCGMs) having fixed number ratio two types functionalized building blocks with 1:1. SCGMs are regarded to a certain degree as size-amplified versions universal synthetic polymers that prepared by precisely connecting molecular nanoparticles (MNPs) polymer chains. Our previous experimental results have shown individually changing one would significantly affect structural parameters crystals, mainly...
To address the limitation and obtain position of drone even when relative poses intrinsics camera are unknown, a visual positioning algorithm based on image retrieval called AGCosPlace, which leverages Transformer architecture to achieve improved performance, is proposed. Our approach involves subjecting feature map backbone an encoding operation that incorporates attention mechanisms, multi-layer perceptron coding, graph network module. This allows for better aggregation context information...
Pavement crack detection is a critical task for insuring road safety. Manual extremely time-consuming. Therefore, an automatic method required to boost this progress. However, it remains challenging due the intensity inhomogeneity of cracks and complexity background, e.g., low contrast with surrounding pavements possible shadows similar intensity. Inspired by recent advances deep learning in computer vision, we propose novel network architecture, named Feature Pyramid Hierarchical Boosting...
Abstract Inspired by biological vision mechanism, event‐based cameras have been developed to capture continuous object motion and detect brightness changes independently asynchronously, which overcome the limitations of traditional frame‐based cameras. Complementarily, spiking neural networks (SNNs) offer asynchronous computations exploit inherent sparseness spatio‐temporal events. Notably, pixel‐wise optical flow estimations calculate positions relationships objects in adjacent frames;...
Recent advancements in autonomous driving, augmented reality, robotics, and embodied intelligence have necessitated 3D perception algorithms. However, current methods, particularly small models, struggle with processing logical reasoning, question-answering, handling open scenario categories. On the other hand, generative multimodal large language models (MLLMs) excel general capacity but underperform tasks, due to weak spatial local object perception, poor text-based geometric numerical...
Abstract Nanoparticle (NP) arrays, particularly those with plasmonic properties, have diverse applications in electronics, photonics, catalysis, and biosensing, but their precise scalable fabrication remains challenging. In this work, a facile chemical‐based strategy is presented for the of NP patterns using combination soft thermal nanoimprinting template‐directed assembly. The approach enables creation well‐defined arrays single‐particle resolution yields over 99%, covering range sizes...
The automatic arrhythmia classification system has made a significant contribution to reducing the mortality rate of cardiovascular diseases. Although current deep-learning-based models have achieved ideal effects in classification, their performance still needs be further improved due small scale dataset. In this paper, we propose novel self-supervised pre-training method called Segment Origin Prediction (SOP) improve model's performance. We design data reorganization module, which allows...
Direct causality detection is an important and challenging problem in root cause hazard propagation analysis. Several methods provide effective solutions to this for linear relationships. For nonlinear situations, currently only analysis can be conducted, but the direct cannot identified based on process data. In paper, we describe a approach suitable both connections. Based extension of transfer entropy approach, (DTE) concept proposed detect whether there information and/or material flow...
In view of the problems low surface detection efficiency and high error rate in domestic wood skin, a corresponding system is designed to identify four types skin defects such as dead knot, slipknot, hole crack on wood.Based image recognition industrial cameras, can collect data real time quickly.Then recognize classify collected images.The great significance rational utilization wood.The experimental results show that safe reliable, has accuracy.
The event camera, a new bio-inspired vision sensor with low latency and high temporal resolution, has brought great potential demonstrated promising application in machine artificial intelligence. Corner detection is key step of object motion estimation tracking. However, most existing event-based corner detectors, such as G-eHarris Arc*, lead to huge number redundant or wrong corners, cannot strike balance between the accuracy real-time performance, especially complex scenes texture that...