A5101865141- Advanced Vision and Imaging
- Human Motion and Animation
- Geophysical Methods and Applications
- Computer Graphics and Visualization Techniques
- Human Pose and Action Recognition
- Olfactory and Sensory Function Studies
- Biofuel production and bioconversion
- Multisensory perception and integration
- Language, Metaphor, and Cognition
- Microbial Metabolic Engineering and Bioproduction
- Machine Learning in Materials Science
- Membrane Separation Technologies
- 3D Shape Modeling and Analysis
- Interactive and Immersive Displays
- Autonomous Vehicle Technology and Safety
- Precipitation Measurement and Analysis
- Smart Agriculture and AI
- Electromagnetic Compatibility and Noise Suppression
- Pelvic and Acetabular Injuries
- Augmented Reality Applications
- Landslides and related hazards
- Model Reduction and Neural Networks
- Medical Imaging and Analysis
- Gyrotron and Vacuum Electronics Research
- Microwave Imaging and Scattering Analysis
Xinxiang Medical University
2024
Stanford University
2023-2024
Peking University
2023
Peking University Third Hospital
2023
Henan Normal University
2023
Chinese Academy of Social Sciences
2022-2023
Phicomm (China)
2023
University of Essex
2022
Shandong Academy of Sciences
2022
Qilu University of Technology
2022
Acute vertebral fracture is usually caused by low-energy injury with osteoporosis and high-energy trauma. The AOSpine thoracolumbar spine classification system (AO classification) plays an important role in the diagnosis treatment of disease. description fractures according to scheme requires a great deal time energy for radiologists.To design validate multistage deep learning (multistage AO system) automatic detection, localization acute body on computed tomography.The CT images 1,217...
Learned graph neural networks (GNNs) have recently been established as fast and accurate alternatives for principled solvers in simulating the dynamics of physical systems. In many application domains across science engineering, however, we are not only interested a forward simulation but also solving inverse problems with constraints defined by partial differential equation (PDE). Here explore GNNs to solve such PDE-constrained problems. Given sparse set measurements, recovering initial...
Rectal cancer is a common worldwide and lacks effective prognostic markers. The development of markers by computational pathology methods has attracted increasing attention. This paper aims to construct signature from whole slide images for predicting progression-free survival (PFS) rectal through an unsupervised artificial intelligence algorithm. A total 238 patients with two datasets were collected the validation signature. tumor detection model was built transfer learning. Then, on basis...
Simulating the time evolution of physical systems is pivotal in many scientific and engineering problems. An open challenge simulating such their multi-resolution dynamics: a small fraction system extremely dynamic, requires very fine-grained resolution, while majority changing slowly can be modeled by coarser spatial scales. Typical learning-based surrogate models use uniform scale, which needs to resolve finest required scale waste huge compute achieve accuracy. In this work, we introduce...
The repair of infected bone defects remains a clinical challenge. Staphylococcus aureus is common pathogenic micro-organism associated with such infections. Gentamycin (GM) broad spectrum antibiotic that can kill S . in dose-dependent manner. However, the systemic administration antibiotics may lead to drug resistance and gut dysbiosis. In this work, we constructed β-tricalcium phosphate/gelatin composite scaffolds incorporated gentamycin-loaded chitosan microspheres (CMs(GM)-β-TCP/gelatin...
In this paper, we introduce a new task called synesthesia detection, which aims to extract the sensory word of sentence, and predict original synesthetic modalities corresponding word. Synesthesia refers description perceptions in one modality through concepts from other modalities. It involves not only linguistic phenomenon, but also cognitive phenomenon structuring human thought action, makes it become bridge between figurative abstract cognition, thus be helpful understand deep semantics....
We show that global lower bounds to the mode volume of a dielectric resonator can be computed via Lagrangian duality. State-of-the-art designs rely on sharp tips, but such structures appear highly sub-optimal at nanometer-scale feature sizes, and we demonstrate computational inverse design offers orders-of-magnitude possible improvements. Our bound applied for geometries are simultaneously resonant multiple frequencies, high-efficiency nonlinear-optics applications, identify unavoidable...
This paper presents a method to leverage arbitrary neural network architecture for control variates. Control variates are crucial in reducing the variance of Monte Carlo integration, but they hinge on finding function that both correlates with integrand and has known analytical integral. Traditional approaches rely heuristics choose this function, which might not be expressive enough correlate well integrand. Recent research alleviates issue by modeling integrands learnable parametric model,...
Modeling and rendering photorealistic avatars is of crucial importance in many applications. Existing methods that build a 3D avatar from visual observations, however, struggle to reconstruct clothed humans. We introduce PhysAvatar, novel framework combines inverse with physics automatically estimate the shape appearance human multi-view video data along physical parameters fabric their clothes. For this purpose, we adopt mesh-aligned 4D Gaussian technique for spatio-temporal mesh tracking...
Abstract Creating plausible motions for a diverse range of characters is long‐standing goal in computer graphics. Current learning‐based motion synthesis methods rely on large‐scale datasets, which are often difficult if not impossible to acquire. On the other hand, pose data more accessible, since static posed easier create and can even be extracted from images using recent advancements vision. In this paper, we tap into alternative source introduce neural approach through retargeting,...
The pretreatment and saccharification processes are both the bottlenecks of lignocellulose bioconversion. Pretreatment plays an important role in efficiency. In this paper, ethanol-assisted FeCl3 was developed to reduce resistance enhance hydrolysis effect various conditions, such as concentration ethanol, temperature time, on chemical compositions physical structures were investigated by scanning electron microscope (SEM), X-ray diffraction (XRD) Fourier infrared spectrum (FTIR). increase...
Recent developments in wave-based sensor technologies, such as ground penetrating radar (GPR), provide new opportunities for accurate imaging of underground scenes. Given measurements the scattered electromagnetic wavefield, goal is to estimate spatial distribution permittivity However, problems are highly ill-posed, difficult formulate, and computationally expensive. In this paper, we propose a physics-inspired machine learning-based method learn wave-matter interaction under GPR setting....
Synaesthesia refers to the description of perceptions in one sensory modality through concepts from other modalities. It involves not only a linguistic phenomenon, but also cognitive phenomenon structuring human thought and action, which makes understanding it challenging. As means cognition, synaesthesia is rendered by more than modalities, cue stimulus can play an important role expressing it. In addition, many efforts, such as identifying semantic relationship between words Therefore, we...
Creating believable motions for various characters has long been a goal in computer graphics. Current learning-based motion synthesis methods depend on extensive datasets, which are often challenging, if not impossible, to obtain. On the other hand, pose data is more accessible, since static posed easier create and can even be extracted from images using recent advancements vision. In this paper, we utilize alternative source introduce neural approach through retargeting. Our method...
Reliably detecting pedestrian objects is crucial for building advanced intelligent surveillance systems in the future. In natural environments, traditional detection methods are challenged by irregular features of field view, such as varying scales, postures, and positions. Inspired mammalian visual layered structures invariant object recognition mechanisms biological vision systems, a novel feedforward neural network investigated to detect moving objects. The proposed model consists two...