A5100362244- Land Use and Ecosystem Services
- Color perception and design
- Advanced Neural Network Applications
- Flood Risk Assessment and Management
- Neural Networks and Applications
- Optical measurement and interference techniques
- Advanced Vision and Imaging
- Emotion and Mood Recognition
- Handwritten Text Recognition Techniques
- Statistical Methods and Bayesian Inference
- Advanced Image and Video Retrieval Techniques
- Consumer Packaging Perceptions and Trends
- Statistical Methods and Inference
- Virtual Reality Applications and Impacts
- EEG and Brain-Computer Interfaces
- Multisensory perception and integration
- Image Processing Techniques and Applications
- Blind Source Separation Techniques
- Energy Harvesting in Wireless Networks
- Matrix Theory and Algorithms
- 3D Surveying and Cultural Heritage
- Machine Learning and ELM
- Environmental Engineering and Cultural Studies
- Machine Learning and Data Classification
- Explainable Artificial Intelligence (XAI)
Tianjin University
2005-2025
Tencent (China)
2024
Hefei University of Technology
2020-2024
Wuhan University
2024
The Ohio State University
2013-2023
Takeda (United States)
2023
Beijing Institute of Technology
2023
Hebei University
2022
Kuaishou (China)
2020
Zhejiang University
2004-2015
The hydrogen desorption properties of MgH<sub>2</sub> are thermodynamically and kinetically improved by the synergistic addition AlH<sub>3</sub> CeF<sub>3</sub>.
Under China’s rural revitalization strategy, peri-urban villages function as pivotal nodes in urban–rural integration. Existing policy research predominantly emphasizes macro-level land and industrial policies, neglecting their spatial development effects on villages. This study addresses the gap by constructing a quantification framework employing Vector Autoregression (VAR) model to analyze impacts development, focusing Dalian’s main districts from 2004 2023. The results indicate...
Linear regression models are widely used in mental health and related services research. However, the classic linear analysis assumes that data normally distributed, an assumption is not met by obtained many studies. One method of dealing with this problem to use semi-parametric models, which do require be distributed. But quite sensitive outlying observations, so generated estimates unreliable when study includes outliers. In situation, some researchers trim extreme values prior conducting...
Land reclamation is a common strategy for coastal cities to expand their land territories. Although the marine environment degradation caused by has been investigated in numerous studies, further exploration terms of spatial–temporal impacts patterns and uses reclaimed lands on required generate detailed policies urban planning. Two-thirds Macao's territorial space come from ocean since middle 19th century. Taking history Macao 1975 2018 as empirical case, this study quantifies landscape...
Quantization is a popular technique used in Deep Neural Networks (DNN) inference to reduce the size of models and improve overall numerical performance by exploiting native hardware. This paper attempts conduct an elaborate characterization benefits using quantization techniques-mainly FP16/INT8 variants with static dynamic schemes-using MLPerf Edge Inference benchmarking methodology. The study conducted on Intel x86 processors Raspberry Pi device ARM processor. uses number DNN frameworks,...
In addition to the intended functionality of product, its affective properties (or Kansei) have emerged as important evaluation criteria for successful marketing product. Recently, immersive virtual reality systems been suggested an ideal platform analysis evolving design because among other things, natural style interaction they offer when examining this paper, we compare effectiveness three types environments evaluating mobile phones that real. Each environment offers different degrees...
To solve the component segmentation problem caused by sticking and overlapping of parts in incoherent handwritten calligraphy characters, we propose a Chinese character part method based on Faster RCNN. The utilizes advantages RCNN multi-scale small targets to difficult problems segmentation. hierarchical features components were used our proposed identify each layer structure obtain components. Qualitative quantitative calculations test effect method. experimental results demonstrate...
Accurately identifying the boundary of urban clusters is a crucial aspect studying development agglomerations. This process essential for comprehending and optimizing smart compact development. Existing studies often rely on single category data, which can result in coarse identification boundaries, insufficient detail accuracy, slight discrepancies between coverage actual conditions. To accurately identify extent clusters, this study proposes compares results three methods dense areas major...
Estimating causal treatment effect for randomized controlled trials under post‐treatment confounding, that is, noncompliance and informative dropouts, is becoming an important problem in intervention/prevention studies when the exposures are not completely controlled. When confounding present a study, traditional intention‐to‐treat approach could underestimate because of insufficient exposure treatment. In recent two decades, many papers have been published to address such confounders...
Quantization is a popular technique used in Deep Neural Networks (DNN) inference to reduce the size of models and improve overall numerical performance by exploiting native hardware. This paper attempts conduct an elaborate characterization benefits using quantization techniques -- mainly FP16/INT8 variants with static dynamic schemes MLPerf Edge Inference benchmarking methodology. The study conducted on Intel x86 processors Raspberry Pi device ARM processor. uses number DNN frameworks,...
This article proposes a distance-based framework incentivized by the paradigm shift towards feature aggregation for high-dimensional data, which does not rely on sparse-feature assumption or permutation-based inference. Focusing outcomes that preserve information without truncating any features, class of semiparametric regression has been developed, encapsulates multiple sources variables using pairwise between-subject attributes. Further, we propose strategy to address interlocking...
Product design is an iterative process that involves, among other things, evaluation. In addition to the intended functionality of product, its affective properties (or “Kansei”) have emerged as important evaluation criteria for successful marketing product. Affective refer consumers' psychological feelings about a and they can be mapped into perceptual elements possible modification toward higher customer satisfaction. products in partially assessed using near photorealistic graphic...
We present Frankenstein, a diffusion-based framework that can generate semantic-compositional 3D scenes in single pass. Unlike existing methods output single, unified shape, Frankenstein simultaneously generates multiple separated shapes, each corresponding to semantically meaningful part. The scene information is encoded one tri-plane tensor, from which Signed Distance Function (SDF) fields be decoded represent the compositional shapes. During training, an auto-encoder compresses tri-planes...
Wetlands form a crucial component of ecosystems, and wetland restoration serves as an effective strategy for promoting sustainable urban development. Spatial support is essential restoration, meaning that research on spatial planning considerable importance. Existing studies primarily focus the analysis distribution characteristics, with limited exploration relationships. This paper aims to explore potential utilizing both characteristics relationships identify issues, thereby facilitating...
Monocular depth estimation plays a crucial role in 3D recognition and understanding. One key limitation of existing approaches lies their lack structural information exploitation, which leads to inaccurate spatial layout, discontinuous surface, ambiguous boundaries. In this paper, we tackle problem three aspects. First, exploit the relationship visual features, propose structure-aware neural network with attention blocks. These blocks guide global structures or local details across different...