A5100642400- Particle accelerators and beam dynamics
- Particle Accelerators and Free-Electron Lasers
- Cryptography and Data Security
- Video Coding and Compression Technologies
- Ferroelectric and Negative Capacitance Devices
- Caching and Content Delivery
- Hydrocarbon exploration and reservoir analysis
- Advanced Graph Neural Networks
- Education and Work Dynamics
- Magnetic confinement fusion research
- Higher Education and Teaching Methods
- Coding theory and cryptography
- Advanced Vision and Imaging
- Geological and Geophysical Studies
- Metal Alloys Wear and Properties
- Gyrotron and Vacuum Electronics Research
- Advanced Decision-Making Techniques
- Metallurgy and Material Forming
- Graph Theory and Algorithms
- Advanced Data Compression Techniques
- Military Defense Systems Analysis
- Nonlinear Waves and Solitons
- Reservoir Engineering and Simulation Methods
- Structural Analysis and Optimization
- Advanced Measurement and Detection Methods
State Key Laboratory of Oral Diseases
2022-2024
Sichuan University
2022-2024
Anhui University of Technology
2024
City University of Hong Kong
2022-2024
Jinan University
2016
Ordnance Engineering College
2015
Xi'an Shiyou University
2015
Beijing University of Posts and Telecommunications
2014
Academy of Military Transportation
2014
Liaoning University of Traditional Chinese Medicine
2014
Graph Convolutional Networks (GCNs) are pivotal in extracting latent information from graph data across various domains, yet their acceleration on mainstream GPUs is challenged by workload imbalance and memory access irregularity. To address these challenges, we present Accel-GCN, a GPU accelerator architecture for GCNs. The design of Accel-GCN encompasses: (i) lightweight degree sorting stage to group nodes with similar degree; (ii) block-level partition strategy that dynamically adjusts...
The growth of the Machine-Learning-As-A-Service (MLaaS) market has highlighted clients' data privacy and security issues. Private inference (PI) techniques using cryptographic primitives offer a solution but often have high computation communication costs, particularly with non-linear operators like ReLU. Many attempts to reduce ReLU operations exist, they may need heuristic threshold selection or cause substantial accuracy loss. This work introduces AutoReP, gradient-based approach lessen...
The proliferation of deep learning (DL) has led to the emergence privacy and security concerns. To address these issues, secure Two-party computation (2PC) been proposed as a means enabling privacy-preserving DL computation. However, in practice, 2PC methods often incur high communication overhead, which can impede their use large-scale systems. this challenge, we introduce RRNet, systematic framework that aims jointly reduce overhead MPC comparison protocols accelerate through hardware...
In the acceleration of deep neural network training, GPU has become mainstream platform. GPUs face substantial challenges on GNNs, such as workload imbalance and memory access irregularities, leading to underutilized hardware. Existing solutions PyG, DGL with cuSPARSE, GNNAdvisor frameworks partially address these but traffic is still significant. We argue that drastic performance improvements can only be achieved by vertical optimization algorithm system innovations, rather than treating...
Recent years have witnessed a variety of applications gaming video coding, while how to improve the coding efficiency has been relatively under-explored. The state-of-the-art standard, Versatile Video Coding (VVC), adopts Reference Picture Resampling (RPR) which allows variation frame resolutions in encoding/decoding. great flexibility supported by RPR motivates us develop bidirectional quality impulse based guidance scheme, an effort fully exploit potential coding. design philosophy...
<title>Abstract</title> Infectious bone defects, a significant challenge in clinical settings, often need advanced therapeutic interventions. This study introduces novel treatment system combining photothermal therapy hydrogel with pulsed drug release mechanism. The employs matrix that is sensitive to the thermal fluctuations characteristic of defect sites, enables controlled and localized release. key component this integration mild agents, which are activated within 40-43°C improve...
Abstract The visual professional education of existing colleges and universities can no longer meet the needs development times. In view such a situation, this paper explores integration book creative design using association rule mining. Based on xAPI technology, data collection platform has been designed to provide security, easy scalability, real-time acquisition for text acquisition. Afterward, is preprocessed so that it be used mining in future. support degree, confidence enhancement...
The growth of the Machine-Learning-As-A-Service (MLaaS) market has highlighted clients' data privacy and security issues. Private inference (PI) techniques using cryptographic primitives offer a solution but often have high computation communication costs, particularly with non-linear operators like ReLU. Many attempts to reduce ReLU operations exist, they may need heuristic threshold selection or cause substantial accuracy loss. This work introduces AutoReP, gradient-based approach lessen...
Graph Convolutional Networks (GCNs) are pivotal in extracting latent information from graph data across various domains, yet their acceleration on mainstream GPUs is challenged by workload imbalance and memory access irregularity. To address these challenges, we present Accel-GCN, a GPU accelerator architecture for GCNs. The design of Accel-GCN encompasses: (i) lightweight degree sorting stage to group nodes with similar degree; (ii) block-level partition strategy that dynamically adjusts...
Video transcoding, which is a long-standing problem in video coding, can be used for variety of applications including bit rate conversion, format conversion and resolution conversion. In this paper, we present an effective transcoding solution from the popular previous generation coding standard, High Efficiency Coding standard (HEVC), to Versatile (VVC) latest developed by JVET. VVC enjoys much better efficiency compared its ancestors at expense higher encoding complexity. To facilitate...
AbstractThe insulin-like androgenic gland hormone is a crucial sexual regulator that involved in the masculine differentiation of crustaceans. As an peptide, has been proposed to act through insulin receptor-mediated pathway. The present study cloned and characterized two receptors (
Waste lead-acid batteries are a kind of hazardous waste, and China attaches great importance to the pollution control their recycling treatment. In this paper, we use historical data installed power generation capacity which has strong correlation with waste grid enterprises, select several modeling methods meet characteristics these data, model analyze respectively, compare prediction values methods, method better accuracy, establish get model, consider factors affecting modify model. The...
Interpolation filters for motion compensation play an important role in video coding. Until recently, interpolation luma (MC) draw most of the attention while chroma compen-sation are less noticeable. This paper comprehensively analyzes different types and presents 6-tap DCT based MC. The proposed designed considering multiple factors, comprising characteristics component, frequency response filter expenses computational complexity. experimental results show method can achieve 0.05%, 1.23%...
This article explores the heterogeneity firm’s internationalization path selection by The New Trade Theory. Considering heterogeneity, based on listed automobile company, do empirical study three which are Export, Foreign Direct Investment(FDI) and Outsourcing, considered five micro factors including productivity, R&amp;D investment other two macro factors, think of scale as explanatory variables, take regression weighted least square, principal components estimation, Panel Data...