Blog

The winners of the 2017 IMAGINE Challenge

by Philipp Schrögel

Posted on 19.09.2017

For our 2017 Open Research Challenge IMAGINE, we asked researches around the globe to submit a walking-visualization-tool of gait sequences based on given inertial sensor-based gait recordings. An interesting task which led to several interesting submission! Thank you to everybody, who participated, you all put many thoughts and effort into it!

(image: pixabay, CC0)

However, we had the difficult task to decide on the question, which submissions/teams stood out of the others and are going to be invited to an Autumn School in Erlangen. Our experts from the Pattern Recognition Lab at the Department for Computer Sciences at FAU and the Department for Molecular Neurology of the University Hospital took on this task and we are proud to present the three selected teams:

  • A submission from the United Kingdom: Dr. Benny Lo, Senior Lecturer at Imperial College London together with the PhD-student Yingnan Sun
  • A submission from Brasil: Malte Ollenschläger and Vinicius Facco, both Doctoral candidates at the Universidade do Vale do Rio dos Sinos (Unisinos)

  • A submission from Switzerland: A team around Cléo Moulin from the young company GaitUp, a spin-off of the University Hospital of Lausanne (CHUV) and the Swiss Institute of Technology of Lausanne (EPFL)

Congratulations! You are going to learn more about the three teams, as we are going to publish some more information on them and their solutions over the course of the next weeks! 

Thank you for the submissions to the IMAGINE Challenge

by Philipp Schrögel

Posted on 14.08.2017

An impression from the ORC Office - the printer is working: the submitted proposals for the IMAGINE challenge are now collected and under review by the experts. Thank you all for your contributions - we received six submissions. This is great, we are looking forward to have a look at them. We will get back to you as soon as possible. Due to the extended deadline, also the schedule for the review probably has to be postponed a little. We will keep you informed!

Deadline extended!

by Philipp Schrögel

Posted on 06.07.2017

The Deadline for our current challenge IMAGINE is approaching with large steps. Thank you all for your interest and questions regarding the challenge so far!

(image: pixabay, CC0)

We also have received some request asking for a bit more time – as a result, we are happy to announce that the deadline for submitting a proposal for the IMAGINE challenge has been moved:

New deadline for the IMAGINE challenge: July 28th, 2017 

We hope you all can make good use of the additional two weeks and work on your ideas and solutions. If you have any more questions, please contact us any time!

Award for the Open Research Challenge!

by Philipp Schrögel

Posted on 05.02.2017

Great news: the FAU Open Research Challenge has been awarded the  "Hochschulperle des Monats" (roughly translated: higher education gem of the month) by the German foundation "Stifterverband". The jury says: "A wonderful project, which is bringing the topic "cooperation" to life in an exemplary way. It brings together researchers and practitioners from all over the world, to jointly address societally relevant scientific questions and also builds new networks and strengthens the global visibility of the university". Thanks to all who worked with us on the project and all who participated! The official announcement ist available here in German: www.stifterverband.org/hochschulperle

The "Hochschulperle" is awarded to creative and innovative projects in the field of research, education and innovation. Each year, different types of projects are considered. In 2017, the focus of the awards lies on "cooperative universities", looking at projects bringing together research institutions and external partners in creative new ways and thereby delivering new inputs for resarch and teaching. The monthly award is chosen by an expert jury. At the end of the year, a public voting on the twelve awards decides on an overall winner as "Hochschulperle des Jahres".

Stifterverband is a joint initiative started by companies and foundations – the only one in Germany to be devoted entirely to consulting, networking and promoting improvements in the fields of education, science and innovation.

Are you ready for the challenge?

by Philipp Schrögel

Posted on 30.01.2017

You have read our anouncement of the next challenge? You are an engineer, doctor or even a full team including doctors and engineers? You already started thinking about the Open Research Challenge and impatiently wait to get started? Well, then we have good news for you! The new year has started, and with it the start of the competition for the challenge is about to arrive soon!

We are making last preparations and finalize the discussions within in our team on the remaining details of the challenge and then we are good to go! Stay tuned, we will officially launch the challenge within the next weeks.

 (Image: FAU/Harald Sippel)

If you are interested in learning more about the scientific and technical background of the project, we recommend this article by our challengers Jochen Klucken and Bjoern M. Eskofier (with colleagues):

"Current challenges demand a profound restructuration of the global healthcare system. A more efficient system is required to cope with the growing world population and increased life expectancy, which is associated with a marked prevalence of chronic neurological disorders such as Parkinson's disease (PD). One possible approach to meet this demand is a laterally distributed platform such as the Internet of Things (IoT). Real-time motion metrics in PD could be obtained virtually in any scenario by placing lightweight wearable sensors in the patient's clothes and connecting them to a medical database through mobile devices such as cell phones or tablets. Technologies exist to collect huge amounts of patient data not only during regular medical visits but also at home during activities of daily life. These data could be fed into intelligent algorithms to first discriminate relevant threatening conditions, adjust medications based on online obtained physical deficits, and facilitate strategies to modify disease progression. A major impact of this approach lies in its efficiency, by maximizing resources and drastically improving the patient experience. The patient participates actively in disease management via combined objective device- and self-assessment and by sharing information within both medical and peer groups. Here, we review and discuss the existing wearable technologies and the Internet-of-Things concept applied to PD, with an emphasis on how this technological platform may lead to a shift in paradigm in terms of diagnostics and treatment."

The full article can be found here: "An Emerging Era in the Management of Parkinson's Disease: Wearable Technologies and the Internet of Things"  IEEE Journal of Biomedical and Health Informatics ( Volume: 19, Issue: 6, Nov. 2015 )

Impressions from the Autumn School

by Philipp Schrögel

Posted on 02.11.2015

Price awarding ceremony at the long night of sciences. (photo: Georg Pöhlein)In October 2015, the three winning teams of the FAU Open Research Challenge were invited to Erlangen to get to know Germany, the region and the university. During the weeklong autumn school, there was plenty of time for cultural, scientific and social explorations. The week started with an excursion to the "Fränkische Schweiz" (Franconian Switzerland) for a hike on the occasion of the „Day of Open Distilleries“.  The next days included an official welcome by the FAU vice president for international relations, a tour through the university and the towns of Erlangen and Nuremberg, industry visits, time for scientific discussions on the challenges and also a trip to Munich. The final event was the public presentation of the Open Research Challenge during the "Long Night of Sciences" and the official price awarding through FAU chancellor Dr. Sibylle Reichert. After a week packed with a lot of activities, the teams left with many impressions and the plan to stay further in contact. We say thank you for the wonderful days and your great contributions to the FAU Open Research Challenge. Below, you will find some pictures from the autumn school. (photo above: Price awarding ceremony at the long night of sciences. Photographer: Georg Pöhlein)

Hiking

Hiking in the Franconian landscape at the beginning of the autumn school. (photo: Thomas Hoffmann)

Hiking

The excursion was a great opportunity to get to know not only the region, but also the other teams and people from FAU. (photo: Thomas Hoffmann)

Welcome by the vice president

The teams are welcomed by FAU vice-president for international relations - Prof. Dr. Günther Leugering. (photo: Sebastian Hemmer)

Team Optimixtli

Team Optimixti from Mexico is discussing their solution at the Chair of Economics, Discrete Optimization,  Mathematics. (photo: Erich Malter)

Team Tartu

Team Tartu worked on a mathematical decoding of holograms. At the Institute of Photonic Technologies, they also had the opportunity to have a look at the laboratory with experimental setups for creating holograms. (photo: Erich Malter)

team UniSA

Team UniSA is visiting the Applied Forensics Computing Group to discuss the various aspects of digital security. (photo: Erich Malter)

At the Long Night of Sciences, a large biennial science festival in the Erlangen Nuremberg metropolitan area, the Open Research Challenge was introduced and the challenges and solutions were presented to the audience. (photo: Philipp Schrögel)

price awarding

The winning teams were awarded by FAU chancellor Dr. Sibylle Reichert.


[Article updated in February 2016]

We introduce: Team Tartu from Estonia

by Philipp Schrögel

Posted on 21.09.2015

And the third of our winning teams comes from Estonia: Team Tartu. They have mastered our photonic technologies challenge and „demonstrated skills in forging holographic security features of future banknotes“ as their university has stated in a press release announcing the success in the FAU Open Resarch Challenge.

Team Tartu: Andreas Valdmann, Ants Remm and Roland Matt

How do you know each other?

We first met through physics olympiads in high school. Later on, we joined the national physics olympiad committee, where it was our turn to make up new physics problems for the students. After that, we all ended up in the physical optics lab in the University of Tartu.

What knowledge and experience did you bring to the challenge?

We all share the interest of solving challenging physics problems. We also have a solid background in wave optics. Ants is a great programmer, Roland did the literature search and Andreas wrote the report.

What did you think you first heard about the challenge?

The first feelings were similar to solving a tricky logic puzzle. After we had some initial ideas about the solution, we started to think about forming a team and really putting some effort into the challenge. We wanted to work as efficiently as possible.

What was your best experience during the challenge?

We really enjoyed working as a team. Before taking the challenge, most of our achievements have been individual (solving problems at olympiads, writing a thesis etc). It was really great to see that we could get results very quickly, if we divide the tasks between us and work in parallel.

What are you looking forward to the most when you are coming to Erlangen?

We really look forward to meeting new people – both the organizers of the challenges and the other winning teams. It is always interesting to get to know people, who are passionate about what they do.

Something about the challenge: How did you approach the problem?

At first, there was some trial and error to get the initial ideas. After we had the holograms decoded, it was quite clear what to do next—search literature for the best existing methods, write the code, do thorough automated testing with different methods and initial conditions, write a clear report. We first did some brainstorming together and then divided the tasks between us, taking into account our individual skills and expertise.

How long did you work on the challenge? Were there special moments, like huge breakthroughs or moments  when you were thinking about giving up?

We had three or four sessions together, each lasting about half a day. We also did some individual work at home. In total, it might have been about one workweek per person. There were three milestones in the challenge. First was getting the initial idea on how to solve the problem. The second was, when we had all the test results and realized, that we have a working solution. We reached the third milestone, when we completed the report and submitted our solution. We never had a really hard moment to think about giving up. About one week before the end of the challenge, it was time for us to shift gears and start putting everything together.

What did you learn during the challenge?

On the technical side, we were introduced to the topic of computer generated holography and the challenges it currently holds. This field was quite new to us and thanks to the Open Research Challenge we are now familiar with its underlying concepts and methods. Apart from that, we learned how to work as a team when solving scientific or engineering problems. We are hoping to build on this positive experience to make our own research more effective.

How would you describe your solution to a non-expert? What makes your solution special and probably helped you win?

Light waves bend around small objects, just like water waves bend around the legs of a sunbather standing in knee deep water. This phenomenon is called diffraction and can be observed for example, when shining a laser pointer at a wall through a narrow slit in a piece of opaque film. In this case, a series of light and dark fringes—the so-called diffraction pattern—will appear on the wall instead of a narrow laser line. The simple slit can be replaced with a pattern of transparent and opaque areas in the film. As a result, the diffraction pattern will also change. By cleverly choosing the pattern on the film, it is possible to display something meaningful in the diffraction pattern, for example some letters, numbers or a small picture. The piece of film is then called a hologram. Some holograms are made completely transparent and instead the thickness of the film is varied. Nevertheless, the same underlying principle of diffraction is at work. It is possible to create many different holograms, that have the same diffraction pattern. There are even more holograms that produce a similar, but slightly different diffraction pattern. Computational methods can be used to figure out the best pattern of the hologram for a given diffraction image. We were given some sample holograms that would be on imaginary future bank notes. It was our task to first find out, what messages were hidden in the holograms. After that, we had to compute our own holograms that would produce identical diffraction images. We used statistical methods to make sure that both the given and our own holograms were as similar as possible. We also found a shortcut to reduce the calculation time, but with added risk of getting caught as our holograms were then becoming “too good” in quality.

If you you want to have a deeper look at the solution, you will find the original report submitted by Team Tartu here:

FAU_OpenResearchChallenge_Report_Team_Tartu.pdf

We introduce: Team UniSA from Australia

by Philipp Schrögel

Posted on 19.09.2015

Today we reveal some more information on the winning team of our digital forensics challenge: Team UniSa from Australia. They have decoded the provided data file and found the secrets within. Here they are:

 

Team UniSA

Team UniSA: Quang Do, Ben Martini and Kim-Kwang Raymond Choo

How do you know each other?

We work together at the School of Information Technology and Mathematical Sciences at the University of South Australia. Quang is a PhD student who focuses on mobile device user privacy preservation and forensics research. Ben is a research fellow who has published in cybersecurity and digital forensics. Raymond is a senior lecturer who has published in cybercrime, cybersecurity, and digital forensics. Naturally with our coinciding research interests, we have worked together on a number of projects over the last few years, including this research challenge.

 

What knowledge and experience did you bring to the challenge?

As a team and individually, we have collaborated on a number of papers in digital forensics and security, and are extending our research to other contemporary technologies such as Internet of Things (IoT) devices. The skills obtained from our previous research experience put us in good stead to complete the digital forensics challenge, as it involved analysing a forensic image collected from an IoT surveillance camera device.

 

What did you think when you first heard about the challenge?

We were excited to find a university lead research challenge on the topic of digital forensics as the digital forensics discipline is a relatively new field of research and certainly a research area that has tremendous growth potential. We found the challenge content to be interesting as it deals with a hot topic in many countries.

 

What was your best experience during the challenge?

Our best experiences were when we successfully located and decoded the executable file used by the intruder to obtain the images, and when we successfully recovered the deleted photos from the evidence file. These two events were integral to our solution and the forensic report that we submitted for the challenge.

 

What are you looking forward to the most when you are coming to Erlangen?

Ben: The Long Night of Sciences sounds very interesting, and I am looking forward to attending.
Quang: I look forward to trying the various local cuisines in Erlangen and Nuremburg.
Raymond: Unfortunately I will not be able to take part in this exciting event due to competing speaking engagements. I look forward to collaborating with the FAU teams in the near future.

 

How did you approach the problem?

We followed digital forensic principles in conducting our investigation. After obtaining the evidence file, we conducted a high level investigation on the evidence to determine the data that may be available, and the format in which the data was stored. After the initial investigation, we sought to answer the specific questions that were posed, in as detailed a manner as possible. It is difficult to explain the specifics of this process briefly. However, if you are interested, much of the detail is provided in our forensic report.

 

How long did you work on the challenge? Were there special moments, like huge breakthroughs or moments when you were thinking about giving up?

We worked on the challenge on a casual basis for approximately three weeks. The breakthroughs included recovering the deleted photos and determining the operation of the executable file installed by the intruder. Researchers don’t give up, and for forensic researchers, attention to detail is crucial!

 

What did you learn during the challenge?

While working in digital forensics allows us to learn about a range of new technologies, we don’t often have opportunities to investigate a complete case, as the challenge was designed to simulate. We often communicate the outputs of our research in academic publications. The experience of writing a forensic report, as required by this challenge, provided a good opportunity for us to learn the specifics of forensic report writing.

 

How would you describe your solution to a non-expert?

We managed to forensically recover deleted evidential data and retrace the steps taken by an attacker on the video surveillance device. This allowed us to understand who did what, when and how in the context of the hacked Smart Home System incident. In a real-world situation, the evidential data could be used in a court of law or for the purpose of civil litigation.

 

If you you want to have a deeper look at the solution, you will find the original report submitted by Team UniSA here:

FAU_OpenResearchChallenge_Report_Team_UniSA.pdf

 

We introduce: Team Optimixtli from Mexico

by Philipp Schrögel

Posted on 16.09.2015

It is high time to learn a bit more about the winning teams of our challenges. We start today with Team Optimixtli from Mexico. They aced the discrete optimization challenge. Who are the brilliant minds behind the team? We asked them some questions:

Rodrigo Alexander Castro Campos, Sergio Luis Pérez Pérez, Gualberto Vazquez Casas together with Francisco Javier Zaragoza Martínez.

 

How do you know each other?

We are all at Universidad Autónoma Metropolitana Azcapotzalco, a public university in Mexico City. Gualberto, Rodrigo, and Sergio took several undergraduate classes on Computer Engineering with Francisco and, later on, the three of them developed their undergraduate projects under his guidance. During their undergraduate studies, Gualberto and Sergio participated in national programming contests and were Teaching Assistants of Francisco. Francisco codirected Rodrigo's Master's Thesis in Computer Science and currently directs Gualberto's Master's Thesis as well as Rodrigo's and Sergio's Doctoral's thesis at our Graduate Program in Optimization. All of us organize together the undergraduate programming contests in our institution.

 

What knowledge and experience did you bring to the challenge?

As part of our Graduate Program, we have experience in modeling discrete optimization problems as integer programs. We have recently started to use Gurobi (and its C++ API) as a tool to solve large integer programs, as well as mixed integer programs. We also know several other algorithmic techniques, such as divide and conquer, dynamic programming, linear programming, backtracking, etc. When problems are too difficult, we have designed approximation algorithms and several kinds of heuristics to solve them. We have a lot of experience implementing algorithms and heuristics in efficient languages such as C and C++, but we also have used scripting languages such as PHP and Python.

 

What did you think when you first heard about the challenge?

Rodrigo: The Open Research Challenge was proposing real-world challenging problems, honoring Germany as one of the most advanced countries in the world. It was a great opportunity to both try to solve them and to travel to Germany :)

Sergio: I thought that it was a very good opportunity to solve a real world problem and the prize was very attractive. This would not be easy because the challenge was open around the world and there are a lot of very smart and talented people around the world but, anyway, I love challenges.

Gualberto: It all started with an e-mail from our advisor: "Do you want to go to Germany?" At first we didn't understand what he meant until we read the challenge and it called our attention. We saw it as an opportunity to learn to solve a different kind of problems to those we are used to in our graduate program. And, of course, there was the possibility to go to Germany in case of winning the challenge.

Francisco: I thought that the Open Research Challenge was a good way to try something new at the same time that we could evaluate the kind of knowledge that we are trying to teach to our graduate students. Therefore, I immediately proposed my students to try it out.

 

What was your best experience during the challenge?

Rodrigo: When we were reaching the end of the challenge, we noticed that we obtained pretty good results for a very difficult problem, and that we could probably win!

Gualberto: To have to learn so much, and to put into practice, in so little time.

Sergio: To see the very good collaboration between the team members and how fast they moved onto new ideas for the problem.

Francisco: My best experience was to see me and my students involved in solving a very difficult, real-life problem, even after we had some initial pitfalls, and not giving up.

 

What are you looking forward to the most when you are coming to Erlangen?

Gualberto: To get to know the German culture and their typical food. I also want to know whether what we did could be improved.

Sergio: To meet other researchers in related areas I'm working on and to establish a kind of research collaboration with them.

Rodrigo: To meet the researchers at FAU and to learn about the German culture.

Francisco: To discuss possible research cooperations with FAU and their bussiness partners. Also to discover a part of Germany and German culture that I don't know.

 

Something about the challenge: How did you approach the problem?

All of us are doing research related to problems that can be modeled as integer programs, so it was natural for us to propose something similar. At the end we proposed a mixed integer program with thousands of constraints and variables. This proved useful, but not enough to solve the challenge: due to our limited computing resources, we could only solve half of the instances using Gurobi this way. At that point, we proposed two ways: an exact approach using dynamic programming and approximate approaches using a simpler objective function. The latter idea proved to be very fruitful, giving us almost optimal solutions for the other half of the instances.

 

How long did you work on the challenge? Were there special moments, like huge breakthrougs or moments when you were thinking about giving up?

We started reading the challenge on May 13th and submitted our report and solutions on August 7th. So we worked on it for almost three months. We did have a huge breaktrough: one day we realized that a simplified objective function (which was originally discarded) had the potential of giving us simultaneously good upper and lower bounds to the optimal value, using a very small fraction of the variables. We did not give up, but we can say that after a few weeks of work we realized that we had misunderstood an important part of the problem. In this, the discussion forum was very useful.

 

What did you learn during the challenge?

Living in a country with almost no trains, we had to learn about train schedules and train energy consumption and production. Then we had to learn about the various file formats involved. We also learned how difficult it is to deal with huge problems and not enough computing resources.

 

How would you describe your solution to a non-expert? What makes your solution special and probably helped you win?

The main difficulty was that our model simply had too many variables. This in turn came from the great number of data points in the energy consumption of each train: one per second. Our simplification reduced this number to one data point each few seconds. Some accuracy was lost in the process, but we gained tremendously in the speed with which we could solve the simplification.

 

If you you want to have a deeper look at the solution, you will find the original report submitted by Team Optimixtli here:

FAU_OpenResearchChallenge_Report_TeamOptimixtli.pdf

And the winners are...

by Philipp Schrögel

Posted on 31.08.2015

... selected! The experts from our challenging institutes (the Applied Forensic Computing Group, the Chair of Economics – Discrete Optimization – Mathematics and the Institute of Photonic Technologies) have reviewed all submissions and identified the best solutions. It was a very difficult task, as the overall quality of submissions was outstandingly high. Especially for the discrete optimization and holography challenge, the final decision was a very close call. Therefore, we have decided to also award two honorable mentions for the runners-up to recognize the excellent work.

If you have not participated in the feedback survey, we would like to ask you for five minutes of your time and let us know what you are thinking about our Open Research Challenge. You will find a short (promised - only 15 questions) survey here: https://de.surveymonkey.com/r/openresearchchallenge2015

And now without further delays:

Digital Forensics Challenge

The winner for the digital forensics challenge is Team UniSA from Australia. Congratulations! All three team members are affiliated with the University of South Australia - Quang Do is a doctoral candidate, Ben Martini is a research associate and Kim-Kwang Raymond Choo is a renowned senior lecturer.

Team UniSA


Photonic Technologies Challenge

The award for the photonic technologies challenge goes to Team Tartu from Estonia. Congratulations! Andreas Valdmann is a PhD student at the Institute of Physics at the University of Tartu. Ants Remm and Roland Matt are currently both Master’s students at the ETH Zurich in Switzerland. Team Tartu was the youngest team participating and their win a further proof that innovation is no matter of age.  

Team Tartu

Team Durham is being awarded with a Honourable Mention for its similar excellent submission to the photonic technologies challenge. Congratulations for the close second place! The team consists of three members from the School of Engineering and Computing Sciences at Durham University: Florian Bryce Soulard (PhD Researcher), Richard McWilliam (Research Fellow) and Alan Purvis (Professor in Electronic Engineering). The fourth member is Joshua Cowling, Senior Applications Engineer at IBEX Innovations.

Team Durham


Discrete Optimization Challenge

And last but not least: the winner of the discrete optimization challenge is Team Optimixtli from Mexico. Congratulations! The team consists of three members of the Graduate Program in Optimization at the Universidad Autónoma Metropolitana Azcapotzalco: Rodrigo Alexander Castro Campos (Doctoral Student), Sergio Luis Pérez Pérez (Doctoral Student) and Gualberto Vazquez Casas (Masters Student) as well as Francisco Javier Zaragoza Martínez, Full Professor at the university.

Team Optimixtli

The optimization challenge was also a close call. So congratulations for the second place and a Honorable Mention go to Team Sodor Diesels. The two team members are PhD student Jørgen Thorlund Haahr and Martin Philip Kidd, PostDoc. Both do their research in management engineering at the Technical University of Denmark.

Team Sodor Diesels