Optimization Algorithm and Data Analysis

We conduct research mainly in the areas of optimization algorithms and data analysis. During these years, we have focused on some specific problems, including online selling, distributed message dissemination, multiple robots collaboration, community discovery and evolution in social networks, online frequency assignment, online code assignment, bin packing, scheduling, etc. We have undertaken 1 NSFC key project, 1 NSFC general project, 1 National Key Research and Development Plan, 1 Key Program of the Chinese Academy of Sciences and several Shenzhen research projects. In these years, we have published more than 100 research papers in some leading journals and conferences.
Ph.D. : Yong Zhang      Vincent Chau


Online Algorithm

Online Merchandise Sale:
A high performance on-line algorithm is proposed for On-line Merchandise Sale, we also prove the lower bounds of performance that any algorithm cannot achieve
Online code allocation:
A high performance on-line algorithm is proposed for code allocation.
Online bin packing:
A high performance on-line algorithm is proposed for two-dimensional bin packing problem.

Intelligence and Automation car parking management system


Research an Intelligence and Automation car parking management system based on the technology of Internet of Things. This system can effectively capture the temporary parking information on the roadside, and manage the parking behavior effectively. We carry out technical research and demonstration application of occupancy detection, data transmission, systems operation, data management and analysis platform.

Spatio-temporal data processing

Based on the spatiotemporal data collected by increasingly emerging sensors such as mobile phones, social media, and Internet of things as well as the high-performance computing technology, we focus on analyzing human spatiotemporal activity patterns in the information age, modeling spatiotemporal phenomena, and simulating spatiotemporal processes. Our research can help breakthrough the limitation of lacking large-scale human activity data sources for decades, and can offer critical methods and technologies for various application areas such as smart transportation, urban planning, public health, mobile Internet service, and big data service in a more general word.
Ph.D. : Jinxing Hu      Ying Lin      Bing He     Liqun Sun      Yuanjun Guo


Development and Application of State Assessment System for Urban Power Network based on the Big Data analysis

According to the power grid, equipment and environmental information, this project is aiming at the research on dynamic evaluation of urban power grid transmission capacity, state evaluation, failure prediction, operation risk assessment, aid decision-making and related key technologies with the basis of big data analysis. At the meantime, this research is applied in the construction of the state evaluation system for the large urban power grid equipment, to improve the safety operation level of the power grid and equipment.
   

Spatial-temporal Analysis Based On Large-scale Data


Based on the spatiotemporal data collected by increasingly emerging sensors such as mobile phones, social media, and Internet of things as well as the high-performance computing technology, we focus on analyzing human spatiotemporal activity patterns in the information age, modeling spatiotemporal phenomena, and simulating spatiotemporal processes. Our research can help breakthrough the limitation of lacking large-scale human activity data sources for decades, and can offer critical methods and technologies for various application areas such as smart transportation, urban planning, public health, mobile Internet service, and big data service in a more general word.

Multimedia Computing and Visual Signal Processing

The SIAT Video Team have long been engaged in the field of Multimedia Communications and Visual Signal Processing for 2D/3D, VR/AR videos, including video coding, visual signal pre/post-processing, and computational visual perception. We are also pursuing exciting and challenging problems in the innovation areas, such as VR/AR, and AI etc. Our research projects include NSFC (General, Youth), Guangdong Provincial Science and Technology Major Project, Guangdong NSF for Distinguished Young Scholar, Shenzhen Municipal Science and Technology Key Project, Shenzhen-International Collaborative Innovation Programs, etc. We have published more than 80 high quality scientific research on journal and conference, such as IEEE Trans. Image Process., IEEE Trans. Broadcast., IEEE Trans. Circuits Syst. Video Technol., IEEE Trans. Indust. Electronics, IEEE Trans. Indust. Informatics. 60 of them are SCI papers and 20 of them are IEEE Transaction papers. Over 10 CN and US patents have been granted. Moreover, our group has many scientific research awards, such as Second Prize Scientific and Technological Advancement Award from Ministry of Education (MOE), First Prize Zhejiang Provincial Science and Technology Award, First Prize Ningbo Municipal Science and Technology Award.
Ph.D. : Yun Zhang      Liangbing Feng      Na Li


3D video technology, Virtual Reality


3D video processing and compression research has been conducted based on high-performance computing platform and optimization theory, with the goal to solve processing, storage, transmission, video quality assessment and other 3D video related issues. Video coding and visual perception are combined to extract human visual redundancy effectively, thus having improved the efficiency of video compression and enhanced the compression efficiency by 30%. In terms of application, An ultra-high-definition 3D live video and on-demand system have been developed. And until now we have cooperated with enterprises in video, such as Skyworth, Phoenix Media and TEMOBI.

Research Group: http://codec.siat.ac.cn/

Computational Biology and Bioinformatics

Our research concentrates on big data analytics in bioinformatics and numerical simulation of molecular structure and properties. The research projects are mainly concerned with highly scalable genome assembly, metagenome-wide association study,protein structure prediction, protein folding simulation and molecular/medical image processing,etc. Research methods mainly include parallel computing, machine learning, stochastic and deterministic optimization. With the help of supercomputers and parallel computing, we have developed high-scalable gene sequence assembly software (SWAP-Assembler) to improve the efficiency of gene data analysis. Using machine learning and stochastic / deterministic optimization algorithms, we study characteristics of protein structures to improve the structure prediction accuracy for both globular and membrane proteins.
Ph.D. : Yanjie Wei     Yong Zhang      Ling Yin


Highly efficient analysis of genomic data at TB - PB scale

A highly scalable genome assembly algorithm SWAP-Assembler was developed based on an asynchronous graph computing framework. Compared with the state-of-the-art tools such as Abyss, the analysis speed of is the fastest. The software is open sourced and has been downloaded about 500 times worldwide.

Environmental and Meteorological Data Mining

Our group has been working in the research of Environmental and Meteorological Data Mining, Environmental and Meteorological Modelling for several years. Our research interest covers: quantitatively forecast of wind and rainfall induced by landfalling and offshore tropical cyclones; tropical cyclones’ intensity forecast; severe weather forecast, interpretation of numerical models; development of ensemble weather forecast system; global and regional climate change; teleconnection between ENSO and regional climate; ecological remote sensing; urban climate simulation, city ventilation design, urban heat island etc. Our group has undertaken 19 provincial and municipal research projects in the area of severe weather forecast, development of ensemble weather forecast system etc. We have published around 50 journal and conference papers.
Ph.D. : Qinlan Li      Liqun Sun      Xiaoxue Wang


Large-scale meteorological data processing and weather nowcast system:


We have provided real-time meteorological data assimilation, severe weather nowcast, and customized service for Shenzhen Meteorology Bureau (SZMB). The research work of quantitative wind and rainfall estimation due to tropical cyclones has been incorporated into the Tropical Cyclones Integrated Operational Platform of SZMB, and has been frequently used in the meteorological information bulletin and weather consultation during typhoon season.

Medical big data

We mainly focus on the medical big data based computer aided diagnosis system by incorporating optimized deep learning algorithms. The research areas include: (1) medical registration, medical segmentation, multi-source medical-data Fusion and pattern recognition; (2) deep learning optimization for small training samples and tensor based deep learning model. Our group has undertaken 9 research projects, including national natural science foundation of china, NSFC-Shenzhen Robotics foundation, Natural Science Foundation of Guangdong Province and Shenzhen Basic Research Project, etc. We have applied more than 20 chinese patients, 2 PCT patients and 1 American patients. We have published more than 30 papers indexed by SCI and EI.
Ph.D. : Shuqiang Wang      Yong Zhang


A Tensor Learning and DTI based CAD system


Presenting a tensor-based classification for DTI data, which can realize the maximum separability of the target images and can largely reduce the algorithm computation complexity as well as can improve the robustness of algorithm. Consequently, it supports DTI based Intelligent CAD theory evidence and core algorithm.