Munich Diaries #5 – Getting in the rhythm3131 words • 15 minute read
Getting in the rhythm
The third week of the lecture period started on the first day of November. As it was also All Saints’ Day, the only in-person class I had (besides the one-on-one consultation with my seminar topic supervisor), the Q&A session for the Game Physics class got canceled, and the professor decided to start it only next week, on the 8th of November. On the other hand, as I had a lot of material to go through before I could ask for a second consultation, there went my only chance to interact with anyone in an academic setting this week. This meant that as far as university matters were concerned, listening to all the prerecorded classes online, and crunching through all of the assignments alone in my dorm room was my lot this whole week.
Luckily, life (especially while on Erasmus) is not only about studying. Looking on the bright side, this week gave me a lot of flexibility in my schedule and opened many possibilities for finding my own rhythm.
Monday - Heimrat im Aqua
On Monday, the house speakers of Biederstein (who are also student-residents here) organized a “Heimrat” – council meeting for all residents. It took place in Aqua (a big common room surrounded by glass windows). The topic of this occasion was lent by increasing tension between some of the students, leading to the involvement of the organization that owns and operates the dormitory.
Ever since the pandemic broke out, among many other things, it has been quite sad to continuously witness (and experience) organization efforts seemingly turn into dust, followed by usually hopeless efforts to deal with ever-changing aspects, expectations, and regulations; leaving people frustrated after vigorous and long-spanning organization efforts.
The Biederstein Dormitory in Munich has been famous for its supposedly awesome parties in the past. As the epidemic situation with Covid-19 surfaced, these parties of course couldn’t have been organized anymore, which led to a more quiet and laid back in-house lifestyle, which I also joined into last month. (As my room is situated right next to the “underground party scene” in the basement of the dormitory, I could see a bit of a silver lining here.)
After regulations (and also the attitude of people) lightened up regarding the pandemic situation, some people jumped right back into party mode.
I’m not sure whether there is a connection between these two things, or it has just always been like this, but some tension built up between residents that prefer a quiet lifestyle in their room, and those who “like it loud”.
The Studentenwerk München – Munich Student Union (or simply StuWerk) is a students’ support association in Munich. They own and run the mensas, dormitories (and also help with accommodation hunting), give support in financial matters, organize events, and run different kinds of advisory services.
The topic of the discussion on Monday was that some heartfelt anonymous emails complaining about party noises arrived to StuWerk, in response to which they introduced measures, including potentially sending a security guard checking the perimeters random times at night – at the expense of the residents. Many people felt this to be a breach of their personal space and weren’t especially fond of the idea.
As many residents of the dormitory feel strongly about a self-governed way of life, open and honest communication is a must. As the complainant’s identity could not be uncovered, it was on the house speakers, and the Biederstein community as a whole, to facilitate a conversation, where everyone has a voice.
This is the setup that led us to the Heimrat on Monday. Many people gathered in Aqua, to the point where it became jammed full of people. I arrived early to snitch a nice and comfortable place on one of the sofas. The house speakers outlined the main topic of discussion, and then everyone started speaking their minds. Even though the main idea was to reach everyone in the dormitory (which includes many international students not being comfortable speaking – and especially discussing matters of importance in German), everyone spoke German during the meeting. With my mid-level German knowledge (and lots of concentration), I could catch the big picture, and even some of the subtleties of the individual comments, but actually partaking in this (sometimes heated) discussion was out of scope for me this time. I also saw lots of familiar faces (pretty much everyone I wrote about so far in previous posts), but I was the only one attending from our WG1. I remember Brecht (an extremely friendly Belgian guy I had the chance to meet right after I arrived) joining the conversation in English, asking things and respondings to comments. The people usually responded in English, but afterward immediately transitioned back to German.
It was great to listen to all the constructive ideas, and people speaking their minds honestly. Even then, as it is usual, as time drew on, I couldn’t escape the feeling of the discussion going around in circles, reverting to the same points. The discussion started at 20:00, and it was 2 hours in when hunger overtook my best efforts at listening to the German discourse on a topic I just entered a month ago, taking place between people that have been living here for multiple years. I excused myself, went to make a hearty dinner, and then went to sleep not long after. The next day, Dávid told me that it drew on much longer after I left, and concisely explained to me the main points I missed.
The conclusion of the night was an agreement between people that communication and constructive attitude have to be improved, and it has to be made clear to everyone living here, that you can always talk to your peers here, and it is always the best course of action, to speak your mind and aspects. After all, we have a saying in Hungary: “The voice of a mute child couldn’t be heard even by its mother.”
Tuesday and Wednesday – Study, Study, Study
On Tuesday, an online (but at least real-time) lecture on how to write a scientific paper was held for my Data Visualization Seminar. I already had my first meeting last week, and this session only further motivated me to jump into researching my topic. I couldn’t wait to start writing the summary.
On Wednesday, I worked on writing system calls (or at least, wrappers for them) in C and Assembly, for my Systems Programming practical course.
After cooking myself a nice lunch, and eating it in front of this week’s 3D Scanning & Motion Capture lecture, I attended the Deutsch als Fremdsprache – Kommunikationstraining B2/C1 (German as a Foreign Language – Communications training B2/C1) course. It was an online course via Zoom, where the emphasis was put on interactivity and discourse in small groups.
Thursday – Running and Salsa
On Thursday, I went running with two of my neighbors at 8:00 in the morning. The scenery and atmosphere were gorgeous as we ran through the English Garden, and along the river Isar up until and across a bridge-like dam, traversing back on the other side of the river. The pace we hit and the company of Isabelle and Robert was nothing short of perfect, and the run ended up being comfortable all the way through, just under 5km in 30 mins.
After this energizing start to my day, I studied effectively all the way until 18:00, at which time the sister of a fellow Biedersteiner organized a free beginner salsa class in Aqua. She was a wonderful and kind teacher. She taught us some preliminary solo steps and we practiced them to music. The session was a great end to my day, letting go of some stress.
Friday – Studystudystudy and going to the movies
In the morning, I started the day by doing my first 3D Scanning and Motion capture assignment. The task was to make a 3D reconstruction based on a regular 2D image, and the depth value corresponding to each pixel. This kind of data is usually recorded with an RGB-D camera, that along with the usual RGB pixel data, also records the depth value, denoting how far away that point is from the camera.
We were given camera parameters, letting us approximate where each pixel of the image would be in the 3D space. Also, as we were given not a single image, but a video (made up of individual frames one after another), and how the camera moves, each new frame contributed its data to the reconstruction. It was not straightforward at all to backproject the pixel into the 3D scene based on camera parameters, but it was an extremely rewarding experience to see the solution come together.
A light-hearted funsie on depth cameras
I came across a really interesting application last year in the Google AI Blog article, The Technology Behind Cinematic Photos. Paired with really interesting supplementary visualizations, in their blog post they write about how they made still 2D pictures “come alive”, adding (or rather, calculating) camera movement, with a trajectory that does not distort the image. They use the depth data from a picture you take with your phone, that’s available thanks to the stereo vision of the multiple cameras most phones are equipped with nowadays. The different vantage points supplied by a multi-camera setup give you an option to calculate the difference between the two viewpoints, allowing you to calculate how far away from the cameras a given object is.
Gathering depth information can differ from device to device though. For each method, the performance, range, and usability throughout different (lightning) environments also vary. Besides the “passive” method of Stereo and Multi-View capture, there are also “active” methods. A famous example is the Microsoft Kinect sensor, utilizing an infrared projector emitting light rays, captured by an infrared camera. The approach is based on the structured light principle. As the relative position of the two devices, as well as the (in this case, fairly random) pattern emitted by the projector, is known, the depth can be calculated based on the data captured by the infrared camera. (For further details, you can read this paper by Microsoft Research, discussing the technology behind the 1st Kinect).
As another example for gathering depth information, Light Imaging, Detection and Ranging (LIDAR) cameras also emit laser light and measure the time it takes for the photons to travel back to the device. Uber used such a capture method for its first “self-driving” car.
First meeting with my i2dl study group
In my last post, I mentioned that I got to be a part of study groups for different classes. In the afternoon, it was time to meet up with my Introduction to Deep Learning (or simply, i2dl) study group for the very first time. We decided to meet in person on the Garching campus, which is situated just north of Munich. Everyone was really nice, happy to meet, and it was great to talk through the current topic of the class and set some intentions for the coming semester.
Going to the movies
Back on Monday, Dávid and I decided to go to the movies. We dropped in the idea to watch the Eternals (the newest Marvel movie at the time) on Friday into the group After Bowling that formed back during the first week I was in Munich. The five of us met in front of a hip burger place, where we decided to have dinner beforehand. It was refreshing to enjoy the company of these lovely people, and share stories and experiences. As one of them couldn’t come to the movie, the four of us ventured forward to the cinema a couple of blocks away.
Going into the movie felt like entering a proper theater, with cushy chairs, red lights, and nice interior design completing the almost art cinema-like experience, which I’m really fond of.
The movie lasted well until after midnight. Venturing home at this time gave me the chance to experience the cold and foggy Munich evening for the first time – which however discomforting might sound, really had a certain charm to it, or at least I certainly felt so in the company of these amazing people I got to spend my evening with.
On Saturday, I went to the end-of-season gathering of MINT, the Ultimate Frisbee team I found a couple of weeks earlier. It was a huge experience, and everyone had an extremely welcoming attitude. I seriously considered joining the team “officially”, which actually turns out not to be the most straight-forward thing to do.2 After having dinner together, evaluating the team’s last year, electing the team officials, coaches, and leaders for next year as well as setting goals for the upcoming season, the evening turned into a party. It was extremely motivating to be a part of reliving all the international trips and tournaments the team did this last year. The atmosphere was great, and I tremendously enjoyed all the energy and friendliness emanating from every single member. I went home late at night, which still made me one of the first people to leave the party.
After giving it a couple of days of thought, I realized I won’t have time for everything, and have to respect myself enough not to try to squeeze in everything. As fun as it seemed like, this made it the end of my frisbee journey – at least for a good while now.
I woke up full of excitement and energy – it was the day of my first in-person class. I made myself a coffee, had a quick breakfast, and headed off to the Garching campus.
Having my place along the metro line U6 is a great perk. It is the oldest U-Bahn line of the network, going from the south part of the city, through the heart of Munich up north until Garching Forschungszentrum – the still expanding Research Center area of Garching (a city just north of Munich). I enjoy being just about halfway between the two. Going south, it connects me right into the inner city, and into the whole public transportation system. This morning, I took it going north, until Garching Forschungszentrum.
Wanting to arrive on time, I ended up with ample time before the class started, even after wandering around for a bit, trying to figure out where the class was taking place. It was a fairly newly built part of the campus. After checking our Covid-vaccination certificates, they let us into a really beautiful classroom with a nice feel to it. The actual lecture was pre-recorded, and this was made out to be a Q&A session after everyone watched the lecture at home by themselves. I enjoyed that Professor Thürey still went over briefly over the class material, and didn’t just ask us whether we have any questions (although I did come prepared with questions, just in case).
As part of this same course, we had to implement some of the techniques we learned in the class in teams. After the Game Physics lecture, I met up with my project team – for the first time in person after assembling ourselves on the Moodle forum of the class. The first assignment was due next weekend: implementing a mass-spring simulator system. (Often used for many purposes in computer graphics like simulating hair, cloth, and deformable solids.3)
My Game Physics project team was once again an assortment of people from really different countries and diverse backgrounds – just between the four of us, we were ranging from full-time BSc, MSc, as well as exchange students, coming from Albania, UK, China as well as Hungary. We had a nice chat over lunch at the Garching mensa and talked through how to approach the project.
Munich Documentation Center for the History of National Socialism
The last planned event I took part in this week was a visit organized by TUM to the Munich Documentation Center for the History of National Socialism. It focuses on the history and consequences of the Nazi regime, with a focus on the role Munich played as “capital of the movement”.
Without going into all the heartbreaking, disturbing, thought-, and emotion-provoking details of our 2-hour guide, a surprising highlight of the trip was meeting Ying from Taiwan. While talking with her, I was reminded again, how surprisingly little I know about all the fascinating sovereign history and culture of far-away countries, and especially about details that have nothing to do with Europe. The world is huge and full of awesome people and stories I know absolutely nothing about!
A visitor from home
On Thursday, at around 18:48, my heart leaped as I saw a familiar red and black winter coat emerge from the crowd, flying towards me. The next thing I knew, I was hugging Bori tight in the middle of the Munich Central Station. I almost couldn’t believe that one of the most lovely and amazing people from home just arrived to spend 6 days with me in Munich.
Wohngemeninschaft (commonly abbreviated to WG) is a common form of living in Germany – and other places in the world. The best translation is shared apartment. The specifics differ, but the main idea is that people share some of the living space in an apartment (usually a sublet). In Biederstein, 6 people share a kitchen, and the bathroom facilities (consisting of the showers, toilets, and faucets), with everyone having their own room (that comes with a further tap). ↩︎
In Germany – as I learned – it’s usual for most associations and gyms not to simply have a monthly membership fee, but to be able to join only for a whole calendar year, with unnecessarily complicated further options. As it was the end of the year, it was not clear to me even what my options are. The process might not be as complicated as it seemed at the time, but it certainly was not straightforward. (Also, having to do all of this on paper – in German, with also the necessity to pay via an automatized yearly withdrawal for one’s German bank account didn’t make it easier.) Anyways, as the last point, I’d like to point out that the team was extremely considerate and helpful, advising every step of the administration process. ↩︎
As an interesting example, Pixar used a mass-spring system-based technique to simulate the curly hair of Merida in their movie Brave. In their technical memo Artistic Simulation of Curly Hair, they outline the method they used. Also, in the Simulation chapter of their course Pixar in a Box on Khan Academy, they wrap it up in an easy-to-follow manner. Originally, this very content inspired me to choose hair simulation as my first computer graphics-related project. ↩︎