Home » Keynote Speakers

Keynote Speakers

Prof. M. E. Fayad
San Jose State University, USA
Title Unified Software Architectural on Demand (USA on-Demand) for Intelligent Systems
Abstract The rapid growth of Intelligent Systems, the challenges and technology, coupled with the tightened intelligent software developments time and production, and cost constraints has imposed tremendous pressure on software industry to create new and innovative designs, which respond to rapidly changing business and operating environments. Software industry must invest in building stable architectures that are flexible and can be easily adapted. We refer to these emerging trends of architectures as Unified Software Architectures on Demand (USA on-Demand): which is based on Software Stability Model and Knowledge Maps as they can be self- adaptive, easily-customizable, self-extensible, more personalizable, self-configurable, and self-manageable accordingly to meet the future requirements and changes in the operating environments.
USA on-Demand presents two complete intelligent systems case studies and much more will be discussed during the conference. A good architecture provides the design principles to ensure, a roadmap for that portion of the road which is yet to be built. Self-configurable and self-manageable architectures, refer to architectures that can manage and “self-heal” its properties vigorously during the reconfiguring runtime of the components, connectors, and the underlying infrastructure.
Prof. Kei Eguchi
Fukuoka Institute of Technology, Japan
Title Design of a High Voltage Multiplier for a Non-Thermal Food Processing System Utilizing Underwater Shockwaves
Abstract In the aging society such as Japan, nutritious and fresh processed foods are required for elderly persons and small children. To provide nutritious and fresh processed foods at low cost, non-thermal food processing utilizing an underwater shockwave is one of the most promising methods. By utilizing the underwater shockwave, this method destroys the cell wall and organization of foods without heating. The non-thermal food processing system is mainly composed of a high voltage multiplier, a pressure vessel, a high voltage relay, and a big capacitor. Therefore, a high voltage multiplier is the vital component to realize an efficient non-thermal food processing system. In this talk, a design of high voltage multipliers is introduced for the non-thermal food processing utilizing underwater shockwaves. First, the problem definition of existing voltage multipliers is described. Next, the circuit configuration and operation principle of the proposed voltage multiplier are explained. Then, experimental results are shown to confirm the validity of the non-thermal food processing system using the proposed voltage multiplier. Finally, conclusion and future work are described.
Prof. Md. Atiqur Rahman Ahad
University of Dhaka, Bangladesh
Title Vision-based Activity Recognition: Present Status & Future Challenges
Abstract Vision-based human activity recognition and analysis are very important research areas in computer vision and Human Robot/Machine/Computer Interaction. Over a decade, a good number of methodologies have been proposed in the literature to decipher various challenges regarding action and activity. However, due to various complex dimensions, a number of challenges still remain unexplored. In this keynote speech, various important aspects of human activity recognition and analysis will be covered. The talk will emphasis on interesting and challenging research aspects to explore in future.
Prof. Takashi Kuremoto
Yamaguchi University, Japan
Title Deep Reinforcemnet Learning: A Promised Way to AI
Abstract Artificial Intelligence (AI) is moving towards a new era in recent years. A game software AlphaGo defeated the European champion of professional Go player in October 2015, and the world champion players in March 2016, May 2017. There are more and more auto-driving vehicles presently, auto-driving bus will be adopted during 2020 Tokyo Olympics and Paralympics. It is widely known that this AI boom is sparked by deep learning, which is a new kind of machine learning method using artificial neural networks (ANNs). However, “deep” only indicates the number of layers of ANNs is more than three, and “learning” which means to modify the synaptic connections between neurons of ANNs usually utilizes the conventional supervised learning methods. On the other hand, reinforcement learning, a kind of goal-directed learning method with trial and error algorithm, may be more suitable to realize AI. In this talk, kinds of deep reinforcement learning (DRL) which are state-of-the-art frameworks with deep architectures and RL are reviewed at first, then a time series forecasting model with DRL is introduced in detail.