Explainable Artificial Intelligence in Image Processing - Ing. Ivan CIK

Ivan Cik, MSc.
Pre-Defense of Midterm thesis in AI

April 23,2021

14.00 - Slovak Time zone,
06.00 am Los Angeles time zone
9.00pm - Beijing Time zone
10.00pm - Tokyo Time zone

Broadcast will be recorded

Nowadays, we are increasingly implementing artificial intelligence algorithms into real life. The development of computing power has enabled the use of more complex algorithms that achieve super-performance on various tasks. With the much larger architecture and nonlinearity of models, the human user ceases to understand what is
happening in the model and how inputs are mapped to outputs. This midterm thesis deals with the concept of explainable artificial intelligence, the division of individual state-of-the-art methods and describes the basic techniques used in image processing, the explanation of which is the subject of the dissertation. The midterm
thesis also sets scientific goals along with hypotheses that should lead to an improved understanding of the behavior of the CNN model, support decision-making, and the introduction of similarity metrics for two similar classes. The technological goal is the development of a framework in the field of medicine with its subsequent testing

http://www.cloudai.sk/people-ivancik/

Pre-Defense of Midterm PhD thesis @tuke PhD study

reviewers :

prof. Choi - S.Korea http://sailab.kaist.ac.kr/
prof. Gregor - UNIZA SK https://michalgregor.gitlab.io/
prof. Turini - Pisa, ITALY https://kdd.isti.cnr.it/people/turini-franco

Supervisor : prof. Marian Mach - http://people.tuke.sk/marian.mach/

Join Zoom Meeting
https://us02web.zoom.us/j/81557729625

Meeting ID: 815 5772 9625
Passcode: XAI