Directed by: Prof. Shipeng Li, Prof. Xi Zhu
This Center is devoted to developing Distributed Intelligent Interaction Systems. Distributed Intelligent Interaction Systems, researched by Professor Shipeng Li, aim to address the issues in traditional natural human-computer interactions caused by the limitations of single or centralized devices and single or limited data sources. This project studies the theory and key technologies in distributed intelligent interaction systems, and it will provide theoretical and technical support for enhancing user experiences in intelligent interactions and extending the ubiquity of intelligent interaction systems.
Directed by: Prof. Dapeng Zhang, Prof. Rui Huang
As one of the world leading research unit, this Center mainly focuses on various fields associated with computer vision, including high-level based image processing, low-level based image processing and biometrics, etc.
Referring to the high-level based image processing, the Center aims to achieve face recognition, image classification, image retrieval and person Re-ID, etc. Specifically, it exploits multiple techniques such as deep learning, hashing learning, reinforce learning, and machine learning to achieve the state-of-the-art performances in different applications.
Referring to the low-level based image processing, the Research Center has also done many works on image denoising, image super-resolution, image deblurring and image compression, etc.
Referring to the biometrics, the Center focuses on two aspects: Security Biometrics and Social Biometrics. In detail, the Research Center has designed various systems including the Fingerprints System, Palm-prints System, and Wrist/hand veins System, etc. for person identification. Additionally, the Research Center has also established Facial Beauty System to measure how beautiful a person is.
It has also exploited the computer techniques to measure Traditional Chinese Medicine (TCM). According to the theoretical analysis of TCM, the Research Center has constructed multiple data collection devices for Tongue/Face Diagnosis, Pulse Diagnosis, and Odor diagnosis, etc. Furthermore, many feature extraction and classification algorithms for TCM diagnosis have also been studied.
Directed by: Prof. Jianwei Huang
With the accelerated evolution of information technologies such as the Internet, big data and cloud computing, crowd intelligence has increasingly played an important role in the information environment of the Internet of Everything (IoE), thus profoundly changing the field of artificial intelligence. At present, the research on crowd intelligence is still in the initial stage, and there does not exist a comprehensive theoretical framework that enables breakthroughs in terms of key technologies such as large-scale crowd intelligence decision-making, incentivization, and coordination.
This Research Center will focus on the basic research of crowd intelligence, with the objective of designing effective incentive mechanisms and coordination control schemes for an efficient crowd intelligence system. Specific research topics included (but not limited to): (1) Game theory in AI; (2) Crowdsourcing and crowdsensing; (3) Incentive mechanism design for crowds; (4) Resource sharing in crowds; (5) Multi-agent reinforcement learning; (6) Federated learning
Directed by: Prof. Tin Lun Lam, Prof. Zhenglong Sun
The AIRS Research Center on Intelligent Robots (RCIR) is committed to working with leading international experts to research leading-edge robotic technologies to give impetus to the application of the technologies in new fields, and popularizing the knowledge of robotics and AI to the society.
- Soft robotics: New sensor and actuator, Micro robotics
- Modular robotics: Self-reconfigurable robotics, Swarm Intelligence
- Multi-robot systems: Distributed computing, Knowledge sharing
- Human-robot collaboration: Cross-scene and behavioral cognition, Affective computing, Medical robot
Directed by: Prof. Kai Hwang
This Center aims to integrate high-performance computing, cloud computing, machine learning and Internet of Things technologies to build a large-scale intelligent industrial data center, so as to promote the integration, application and development of industrial intelligent cloud and industrial Internet of Things technologies over the Greater Bay Area.
There are 8 highlights of major research tasks of the center, including: (1) Cognitive computing and smart industrial cloud; (2) Smart cloud system architecture and virtualization technology; (3) Distributed smart cloud storage design and industrial database system construction; (4) Sensing system of Internet of Things; (5) Optimization of industrial supply chain framework; (6)Machine learning modeling and algorithm selection; (7) Application of data analysis tools; (8) Industrial application of deep learning
Directed by: Prof. Hongyuan Zha
Research interests of the Center cover a broad spectrum of foundational machine learning theories and algorithms as well as applications areas with societal impacts. The Research Center’s key focuses include:
(1) statistical and mathematical theories of machine learning, machine learning methodologies based on optimal transport theory, stochastic differential equations for MCMC, partial differential equations for generative models, and mean field game theory; (2) deep reinforcement learning and multi-agent reinforcement learning, off-policy policy evaluation, and robust MDP; (3) differential privacy and private learning, statistical utility and privacy tradeoff, and constrained distributional robust learning; (4) point process models and event history analysis, neural intensity functions and random measures, co-evolution of dynamics on the networks and network topology changes; (5) machine learning and reinforcement learning for combinatorial optimization problems, graph neural networks, generative models for dynamic graphs; (6) zero- and few-shot learning and meta-learning. Applications include machine learning for scientific computing, social networks, transportation and healthcare.
Directed by: Dr. Ning Ding
Research Center on Special Robots is committed to the research and development of special robots for large-scale city infrastructures including bridges, tunnels, transmission and underground pipe network, and to provide a life-cycle management that aids or replaces human in the operation, maintenance, overhaul and retirement of equipment and facilities. It mainly focuses on two research subjects: (1) design and motion control of bio-inspired special robots for infrastructure maintenance tasks; (2) nondestructive inspection for key components of infrastructures and few-shot learning based recognition of typical diseases.
The first research subject involves topics such as structure design and motion control of high-load high-maneuverability bionic climbing robots subjected to constrains of environmental space and energy, and optimization of robot structural parameters and motion strategy based on deep learning. The second research subject aims to solve the problems of contact/non-contact detection of surface/internal diseases and damages for infrastructures, and typical disease recognition based on few-shot learning.
Directed by: Dr. Yongquan Chen
The mission of this Center is to create robots and intelligent systems that are able to autonomously operate in urban complex and diverse environments. Its research interests include novel methods for perception, abstraction, mapping, path planning, manipulation and cooperation between multiple agents. This Center is well equipped with experimental and computational facilities. Students and staffs work closely together to conduct research in the discovery, design, and implementation of novel applications of urban unmanned systems.
Directed by: Dr. Tian Dihong
Extended Reality, including virtual reality (VR), augmented reality (AR), and mixed reality (MR), is expected to be the next form of human-computer interaction and personal computing, just as smart phones have done in the past decade. At AIRS, the Research Center on Extended Reality (CXR) focuses on conducting research and advanced development of key technologies for extended reality, including computational imaging, display, and automated content generation using artificial intelligence (AI) technologies. The Center is also developing pilot systems to demonstrate applications of XR in manufacturing, collaboration, and education.
The Center is directed by Dr. Dihong Tian who is IEEE Senior Member, and was formerly technical leaders at Cisco System and Samsung in the USA.