Academic roles

My career as an official started in 2010 when I was persuaded to become a member of the Disciplinary Committee. I served on it for six years (i.e., all three two-year terms allowed by the Disciplinary Regulations).

In March 2011, I first entered the Academic Senate of the Faculty of Informatics as a replacement for Michal Kozubek, who resigned due to his candidacy for the dean position. At the end of 2011, elections for academic senates were held, in which I was elected to both the small faculty senate and the large university senate. In January 2012, to my great surprise, I was elected as the chairman of the faculty senate in a closely contested battle with a heavyweight, Jiří Zlatuška, by a narrow margin of 4:5 during the inaugural meeting of the new faculty senate. This was likely influenced by the fact that I held the second strongest mandate from the academic community (after the widely respected Mojmír Křetínský, who was then the department head and did not wish to take on the role of senate chairman), and I was perceived as a less confrontational candidate for the chairmanship by some of the senators.

In the following term starting from 2015, I was re-elected to both academic senates (the small and large ones), but I no longer wanted to serve as the chairman of the small senate. In the spring of 2015, in a repeated and exciting election, we chose Jiři Zlatuška as the (old) new dean. I considered the unexpected offer to become the vice-dean for bachelor's and master's studies from September 2015 for a while. In the end, I decided to accept the new challenge. I was put in charge of the admissions procedure, the organization of the state final exams and the handling of student matters (fees, scholarships, graduation ceremonies, deciding on student applications, etc.)

After his re-election in 2019, Dean Zlatuška offered all vice-deans to continue leading the faculty. After four years, I knew that I was not sufficiently dedicated to all components of academic work while serving in a demanding position. In particular, my scientific performance was declining. Nevertheless, after careful consideration, I decided to continue in the role of vice-dean for one more term. Several factors influenced my decision: (1) Dean Zlatuška always supported me in critical moments and helped me when I needed it, (2) I respect him as a person with great insight and perspective, and although some of his ideas or suggestions seemed crazy to me at first, I always found a reasonable and meaningful core in them that enriched me internally, (3) the Faculty of Informatics has a fantastic Study Department that I have come to love, (4) colleagues I respect have encouraged me that I am doing a good job as Vice Dean and that the second term will be easier. It probably could have been like this if the COVID epidemic has not come in 2020 and turned an easier term into a more challenging term. I am grateful that in an unexpected and turbulent time, I have already had at least 4 years of experience as a Vice Dean and capable colleagues in the faculty and at the head of the university.

My second term as Vice Dean ends in August 2023. I look forward to taking a break from these activities (at least for a while) after more than twenty years in various academic roles. I view my administrative roles as a service to the university, which is important, but I do not consider it the main purpose of my life. I have learnt a lot in my academic roles, and especially my time as a Vice Dean has significantly influenced my view of the world. As Vice Dean for study affairs, I had to look much more deeply into the lives of students and learn about the real problems that affected their studies in order to make responsible decisions about their fate. This was the most valuable experience for me.

Teaching

Since 2005 I have been teaching courses related to digital image processing and analysis. For the longest time I have been teaching the PA166 Advanced Methods of Digital Image Processing course, which gives an overview of image processing methods that are formulated as solutions of partial differential equations (e.g. diffusion filtering, active contours) or as minimization problems of the calculus of variations (variational filters, Mumford-Shah and Chan-Vese segmentation, variational optical flow, etc.). I got a closer look at the world of mathematically well-founded methods during a month-long stay in Joachim Weickert's MIA (Mathematical Image Analysis) group. Being around this first-class scientist (among others, winner of the 2010 Gottfried Wilhelm Leibniz Prize) and a great teacher, who is also very friendly, funny, and extremely hard-working, was an unforgettable and very enriching experience for me.

The second course, which I have been teaching for a similar length of time, but since 2011 with a lower frequency (once every two years because it was less attended), is PA170 Digital Geometry. The course content is largely based on the book [Klette, Rosenfeld, 2004] and is only compulsory for students in the image analysis and processing specialization. I have taught other courses for a shorter period of time. For two years, when my colleague and also my brother Petr was abroad, I taught his course PA173 Mathematical Morphology. From 2006 till 2012, I taught the more general course PV131 Digital Image Processing, taught by Michal Kozubek. I have supervised dozens of students in the special project course PV162 Project in Image Processing. I write more about the individual guidance of students below.

In 2013, I was tasked with redesigning an important compulsory course for Computer Graphics students (from 2019, the entire Visual Computing study programme) PV189 Mathematics for Computer Graphics, which was not well received by students. It covers the mathematics needed in computer graphics, i.e. it mainly discusses linear algebra, quaternions, interpolation, model fitting, and calculating geometric properties of curves and surfaces. My colleague Dima Sorokin, a graduate of applied mathematics at Lomonosov University in Moscow, who was then a postdoc with us, helped me a lot with the change. We rearranged and supplemented the content and prepared a set of exercises that students must solve during the semester. I have been teaching this course ever since and I am constantly trying to improve it, to make the explanation as clear and illustrative as possible.

All of the above mentioned courses are part of the master's degree programme in Visual Informatics, which has been open since the autumn semester 2019.

When I started as Vice Dean, I wanted to get to know the first-year bachelor's students and learn about the knowledge they brought from secondary schools. I therefore signed up to be on the front line and became involved in teaching their compulsory courses as a seminar tutor. Initially (since 2016), I taught two groups of IB000 Mathematical Foundations of Computer Science, which covers the basic mathematics required for studying computer science, and the exercises involve solving mathematical problems at the blackboard. After two years, I replaced one group of students in IB000 with one group of students in IB111 Foundations of Programming, which covers the basics of algorithmization and students program simple algorithms in exercises. My motivation at the time was that I wanted to learn another introductory course and force myself to learn Python, in which the teaching took place. Until then, I had routinely programmed mostly in C++ and felt that I didn't want to live without Python anymore :-).

Before the start of the autumn semester in 2022, I was offered the opportunity to join the DUCIT Teaching Lab, a course designed for beginning student tutors to develop the competencies needed to lead beneficial exercises. In the past, I had acquired my teaching skills more through practice than through studying pedagogy, so I was unsure whether to accept the offer and whether I was the right person. In the end, I nodded and replaced may involvment in teaching the basics of programming with participation in the Teaching Lab. I don't regret the decision. I became part of an experienced, friendly, creative and innovative team of people where I could provide my expertise and learn a lot from colleagues with a stronger theoretical pedagogical background. However, two semesters were enough for me to realize that teaching soft skills is not what I want to do in the long term.

Research and development

My research and development activities are primarily associated with my work for the Centre for Biomedical Image Analysis (formerly the Laboratory of Optical Microscopy), which was founded by Michal Kozubek at the beginning of my PhD studies in 1998. Together with my brother Petr, we were Michal's first PhD students. We helped him with the development of the so-called image cytometry for the purpose of research on the spatial organization of chromatin in cell nuclei, which was then carried out at the Institute of Biophysics of the Academy of Sciences. The core of the cytometer was an automated confocal fluorescence microscope that, together with appropriate software we developed, captured and analysed 3D digital images.

During this period, I learned a lot about optical microscopy (especially fluorescence and confocal) from Michal, and from the (molecular) biologists I was in regular contact with about cell biology (and especially the cell nucleus) and the FISH (Fluorescence In Situ Hybridization) technique, which was used to prepare samples for spatial imaging on a confocal microscope. However, I also gained insight into other techniques, including the use of GFP (Green Fluorescence Protein) to investigate dynamic processes in living cells. My PhD thesis focused on three-dimensional reconstructions of objects for image cytometry.

I participated in the creation and development of the I3D library (written in C++) for loading and processing 3D images. On top of it we developed Viewer3D for viewing 3D digital images. For image acquisition on automated fluorescence confocal microscopes and spatial analysis of cells (cell nuclei), the complex software Acquiarium was developed under my leadership and ran for several years on several microscopes in Brno. As soon as microscopy companies started to supply more general tools to solve the problems of our biologists, we stopped the development.

After finishing my PhD studies, I was involved in several projects (mainly I worked closely with the charming, exceptional woman, scientist and excellent manager Eva Bártová), in which I was tasked with the analysis of image data from fluorescence microscopes (multichannel, three-dimensional and time-varying data). Thanks to the intensive collaboration with biologists, I got to interesting problems for which we were looking for general solutions using the latest informatics approaches. The problems I solved most often were segmentation and object tracking.

Together with my PhD students I investigated the applicability of modern image processing approaches for fluorescence microscopy. With Jan Hubený I worked on methods based on partial differential equations and variational optical flow, with Martin Maška on fast implementations of level-set methods, with Ondřej Daněk on the use of minimization by graph-cuts, and with Karel Štěpka we investigated algorithms for small object detection. With postdoc Dmitry Sorokin, whom I mentored, we worked on global cell motion compensation.

During the preparation of the first edition of the Cell Tracking Challenge in 2012, I was responsible for and involved in developing a suitable measure for evaluating the performance of cell tracking algorithms.

As part of an interdisciplinary research project with the Department of Experimental Biology, Faculty of Science and the Department of Histology and Embryology, Faculty of Medicine (both at MU), we have been working on the analysis of exosomes and developed the TEM ExosomeAnalyzer.

Since 2011, CBIA has been involved in industrial cooperation with three companies, with whom we cooperate intensively, among other things, on the level of solving bachelor and master theses, which are usually supervised by me. With Tescan we are working on tasks related to improving image capture, processing and analysis in the field of Scanning Electron Microscopy (SEM), with Sanezoo we are working on tasks related to visual inspection in industry and object reconstruction using cameras for robotic applications, and finally with Goalsport we are working on image analysis problems in the sports environment.

Popularization of science (and mathematics)

I enjoy finding ways to explain concepts, ideas or principles in a simple and appealing way. It's probably related to my biggest hobby (puzzle hunt games), where it's basically the same thing. Providing the other person with only and only the information that he/she knows for sure and that he/she necessarily needs to discover a certain (unknown) principle. Everything should be presented in an entertaining way and constructed in such a way that correct understanding is the most likely option when piecing the information together. The only difference is that in popularization, information is visualized or demonstrated as clearly as possible (especially a moving picture is worth a thousand words), but in puzzles, information is hidden so that it cannot be seen at first glance. When players eventually discover it during the solving process, they know it's not a coincidence and have to work with it seriously.

I have always enjoyed being involved in events like Researchers' Night or going to high schools to give talks on image analysis. In a lecture entitled (deliberately with a long subtitle) "Digital Image Analysis, or Can Criminalists Truly Extract the Face of the Murder from the Reflection in the Victim's Glasses Using Image Analysis of Surveillance Footage from an Industrial Camera at the Crime Scene?" that I gave at several high schools, I tried to demonstrate in an accessible way the image formation and the theoretical limits of extracting information from a (digital) image. I owe the ideas for some of the physics demonstrations to a wonderful and warm-hearted person, Eduard Schmidt, who gave me one semester of individual tutoring while I was preparing for the entrance exams for my PhD in physics. I will never forget sitting in his office at the Rector's Office, and he took off his glasses, took out a white paper and demonstrated the formation of an image. I was equally amazed at his demonstration of the reddening of the sun at sunset in his lecture on light scattering in his semester course on optics for professional physics students, which I later took out of interest.

During a holiday in Ireland in 2015, my family and I visited the Science Gallery at Trinity College Dublin. I was absolutely amazed by the exhibition there, which demonstrated the image digitisation. I was so excited by this great idea that later at home I constructed a smaller replica out of wrapping paper, wood and cardboard that I use in my Digital Geometry classes and CBIA lab at various popularization events like Researchers' Night.

I am fascinated by websites that explain mathematics in a somewhat playful and visually engaging way (e.g. 3Blue1Brown, Numberphile, Brilliant, Seeing theory, Wolfram demonstrations, Geogebra, Khan Academy, or Desmos). Under this influence, I have started working on study material for Visual Computing students (and not only for them) to help correct the frequent misunderstandings of basic mathematical principles and concepts that I encounter in oral exams and in my students' teaching.

Master's and Bachelor's theses (projects)

Individual work with students is probably the most enjoyable of my teaching activities. During one-on-one contact, one can respond to the needs of the student in a non-violent and completely natural way. Moreover, the mutual cooperation is also beneficial for me. I have learned something new and interesting from many students. Especially from the excellent ones, who are a pleasure to work with. Many of them have won various awards for their work (most often the Dean's Award for an outstanding final thesis), which always makes me happy.

An overview of the free topics and the students I have supervised so far can be found in IS. I also try to offer interesting topics in the project course PV162 Project in Digital Image Processing, which runs every semester. Most of the topics can be developed into a bachelor or master thesis. I am also happy to guide students who want to tackle their own topic (if their topic intrigues me).