WUHAN, China (July 20, 2023) - Wuhan Optoelectronics Forum No. 188 was successfully held in Auditorium C117 at Wuhan National Laboratory for Optoelectronics (WNLO) in the morning of July 20th. Prof. Zhonghai Lu from the KTH Royal Institute of Technology delivered an inspiring speech entitled Deep Learning: Rising, Accelerating, and Risking. Prof. Fei Wu, Deputy Party Committee Secretary of WNLO, chaired the forum and awarded Prof. Zhonghai Lu the forum medal.
Professor Zhonghai Lu started by discussing the three significant milestones in the rapid development of deep learning: Go-playing, image recognition, and conversations. These critical applications have brought about significant changes in our lives. He further emphasized that the success of deep learning not only relies on powerful models, algorithms, and the support of large-scale data but also depends on hardware acceleration. He briefly introduced common methods of hardware acceleration and stressed its importance. Without strong computing power, it is impossible to effectively train large and complex models or achieve real-time inference. Hardware acceleration also provides more opportunities and possibilities for related research. Professor Lu also highlighted the potential risks associated with deep learning, such as vulnerability to deception and the lack of interpretability in results. Deep learning is challenging in some scenarios and requires further research. The atmosphere at the report was lively, and the teachers and students of the research center had a lively exchange with Professor Lu on the content of the report.
Professor Zhonghai Lu is a full professor in Electronic Systems Design (with specialization in Dependable and Autonomous Systems) at Division of Electronics and Embedded Systems, Department of Electrical Engineering, School of Electrical Engineering and Computer Science (EECS). He received BSc. from Beijing Normal University in June 1989, and earned MSc. in June 2002 and PhD. in March 2007 from KTH Royal Institute of Technology. After PhD., he conducted postdoctoral research at KTH. In October 2012, he received MBA in Innovation and Growth from the Business and Innovation Development unit, University of Turku, Finland. He has published over 200 papers in the fields of Computer System Architecture, Embedded Systems, AI Hardware, and Applications. Professor Zhonghai Lu serves as the Program Director for the Embedded Systems master program at KTH and is also an Associate Editor for ACM Transactions on Architecture and Code Optimization (TACO).
