Institute of Information Science, Academia Sinica

Artificial intelligence (AI) is currently a popular topic and has gained high expectations for its future. AI research is not new and can be traced back to the 1940s and 1950s. The early focus was on the use of high-level programming languages such as Lisp and Prolog as appled to areas such as "game theory" and "expert system." In 1982 the "Fifth Generation Computer Project" was launched with a focus on AI and parallel processing. However, since 1987, due to bottlenecks in AI research, the artificial neural network (ANN) has flourished. In 1997, IBM Deep Blue defeated the world chess champion was an important milestone. AlphaGO, which emerged in 2015, has upset the world of the Go game. Reaching an estimated level of 14 dan, AlphaGO cannot be beaten by any professional player today. But what is the core technology of AI? Is the success on the Go game can be directly applied to other fields (such as autonomous vehicles)? At present, the so-called AI, besides the continuing exploration of machine learning and knowledge representation and reasoning, has almost completely abandoned the early high-level programming languages. Today's AI has widely absorbed other technologies, including signal processing, image processing, and speech processing, pattern recognition, parameter estimation, robotics, fuzzy logic, and ANN. This presentation will report on some research results of expert system and ANN applied to medical image processing, with the intention to explore its technical limitations including the inefficiency and unpredictability of the expert system; the impact of ANN structure, depth, and the choice of training sets on its generalization. The past can teach us about the future. More substantive and critical assessments can thus be formed on the current and future technologies.

Ying Sun received the BS degree from the National Taiwan University in 1978 and the PhD degree from the Worcester Polytechnic Institute in 1985. Although all his degrees are in Electrical Engineering, he has always been interested in the field of Biomedical Engineering. In 1985 he joined the faculty of the University of Rhode Island, where he is now a full professor in the Department of Electrical, Computer and Biomedical Engineering and the Director of the Biomedical Engineering Program. His research areas include medical devices, biomedical signal processing, medical imaging systems, modeling of physiological systems, neuroengineering, and assistive technology for persons with disabilities. His research has been supported by the National Science Foundation, the National Institutes of Health, hospitals, and medical device industry. His research results have been published in over 250 papers in professional journals and conferences. He is the co-inventors of seven issued patents. He has served as a consultant to medical device industry and has written several professional software packages for cardiovascular research and medical imaging. He visited Linköping University (Sweden) in 1992 and University of Vienna (Austria) in 1993. Since 2010, he has come back to Taiwan every year as a short-term Visiting Professor at the National Sun Yat-Sen University. Ying Sun has played the classical guitar for many years. Recently he and his colleagues from relevant departments have developed a general education course entitled “Technologies and Music.”