Aug 5 – 16, 2024
Europe/Zurich timezone

About the lectures

We will have 4 lectures:

I. Generative Modeling – from probabilistic PCA to Diffusion Models by Prof. Johannes Brandstetter (JKU Linz):

This lecture starts by deductively introducing probabilistic PCA as an example of a latent variable model with closed-form posterior. Next, Variational Autoencoders are derived as non-linear extension thereof, with emphasis on the evidence lower bound and their variational inference characteristic. Finally, diffusion models are introduced as hierarchical VAEs with fixed priors. The technical concepts are relaxed by interactive notebooks.

II. Science Communication by Dr. Sascha Vogel (Science Birds):

The course is offered by Dr. Sascha Vogel.
Sascha is a science communicator and the founder of science birds, a company dedicated to making science accessible and engaging for the public. He holds a background in theoretical physics, worked as a researcher abroad and has more than 15 years of science communication experience. 
In this field he is known for his unique approach to science communication, including his popular "Physics in Hollywood" program, which examines the scientific accuracy of popular films and engages audiences with a blend of education and entertainment. 

The science communication program will provide a comprehensive introduction to the principles and practices of effective science communication. 
Participants will engage in hands-on activities, including the recording of podcasts and the creation of dynamic social media content, aimed at making scientific concepts accessible and engaging to diverse audiences.  In many practical sessions, students will develop essential skills for communicating complex scientific information in clear, compelling, and creative ways.
By the end of the program, students will have produced a variety of science communication outputs and gained the confidence to share their scientific knowledge with the public.  

III: AI4Science by Prof. Kai Zhou (CUHK and FIAS):

Dr. Zhou received his PhD degree in Physics with ‘Wu You Xun’ Honors from Tsinghua University in 2014. Afterwards he went to Goethe University Frankfurt to do postdoctoral research work at the Institute for Theoretical Physics (ITP) . Since August 2017, he is a FIAS Research Fellow focusing on Deep Learning (DL) applications in research, and guides the ‘Deepthinkers’ group at FIAS as their group leader. Dr. Zhou has a very broad interest in physics and AI/DL application in different fields. Recently with his collaborators he developed a deep learning based strategy to help efficiently extracting essential properties of the very involved dynamical evolution from only final
observation, they applied it in Heavy Ion Collisions to construct an Equation-Of-State (EOS) meter which is an ultimate goal for the experiments efforts. Since 2024 he is Assistant Professor in the School of Science and Engineering of CUHK Shenzhen.

IV Computer-aided diagnosis method and system based on video motion analysis by Dr. Riumin Li (Xidian):

Dr. Ruimin Li is working at the Academy of Advanced Interdisciplinary Research, XIDIAN University as HUASHAN Associate Professor. She is a master’s supervisor, IEEE member, and member of Visual Sensing Committee of Chinese Society of Image and Graphics (CSIG). Her main research interests include video motion recognition, action quality assessment, and intelligent aided-diagnosis system design, aiming to provide new methods for assisted diagnosis and rehabilitation training of developmental coordination disorder (DCD). She is presiding over one Youth Program of the National Natural Science Foundation of China (NSFC) and participating in one General Program of NSFC. She has published more than 20 papers in international conferences and journals, i.e., Pattern Recognition, IEEE Transactions on Multimedia, IEEE Transactions on Circuits and Systems for Video Technology, and IEEE Transactions on Neural Systems and Rehabilitation Engineering.