Seminars

The registration for all computer science seminars will open in April.

This system is used to distribute students among the available actual seminars. To register for any of the other seminars that are offered by the computer science department, you have to register here until April 10th 23:59 CEST. You can select which seminar you would like to take, and will then be automatically assigned to one of them on April 12th. Please note that the assignment cannot be optimal for all students if you drop the assigned seminar.

Please note the following:

We aim to provide a fair mapping that respects your wishes, but at the same time also respects the preferences of your fellow students.
Experience has shown that particular seminars are more popular than others, yet these seminars cannot fit all students. We thus encourage students to select their preferences for all available seminars, which eases the process to assign students that do not fit the overly popular seminars to another, less crowded one. Each student must therefore select at least two seminars (with priority from "High" to "Low").
If you are really dedicated to one particular seminar, and you do not want any other seminar, please select the "No seminar" as second positive option. However, this may ultimately lead to the situation that you are not assigned to any seminar. Also, choosing "No seminar" as second option does not increase your chances of getting your first choice.

The assignment will be automatically performed by a constraint solver on April 12th, 2019. You will be added to the respective seminars automatically and be notified about this shortly thereafter.

Accountability in Theory and Practice by Robert Künnemann

Accountability is the ability of actors within a system to hold others accountable, e.g., for violating some policy or for acting in bad faith. Accountability is used to describe desiderata in social sciences, be it the accountability of a government or the accountability of actors in a company context, but we look at it from through the lense of security.

Aim of the seminar: understand accountability in different contexts and learn to use it as a guiding principle in system design; overview about concrete proposals for accountable security mechanisms. Last but not least: argue a point in reviews and rebuttals.

Modus operandi: Students bid for "authorship" of two papers and review another three ( at least one theoretical and one practical). The "authors" respond to the rebuttals (in groups). At the end, each student presents "their" paper.

Requirements: Basics on security: either you've attended a course like Security or Cryptography course, or you can argue that you are knowledgeable in information security. In the latter case, please use the motivation field to make this argument, e.g.: "I have not attended any security-focused course, but I earn my living hunting vulnerabilities for money."

Places: 12

Advanced Topics in Modern Cryptography by Kamil Kluczniak

Fully-Homomorphic Encryption (FHE) schemes and Multi-Party Computation (MPC) are fundamental tools in modern cryptography. For decades FHE and MPC schemes have been abstract concepts living in the realm of cryptographic theory. In recent years those systems have seen major improvements in terms of efficiency and practicality. In short, FHE and MPC schemes have become practical enough to be considered for applications in private delegation of machine learning models, and applications to privacy-preserving distributed Genome-wide association studies. This seminar is concerned with the fundamentals and practical aspects of FHE and MPC. We will heavily focus on the general understanding of the presented schemes, their implementation, and concrete efficiency. By the end of the seminar participants should possess fundamental knowledge about FHE and MPC and should know the state-of-the-art libraries and developer tools which are nowadays available. Among others, we will discuss cryptosystems and their implementations which are part of
- FHEW/TFHE (FHE library)
- HElib (IBM)
- SEAL (Microsoft)
- SPDZ, Mascot, Overdrive (MPC)
- Other open source tools

Seminar web page: https://cms.cispa.saarland/atmcrypto/

Requirements: Students are required to have basic knowledge of Linear and Abstract Algebra, Cryptography (Fundamentals), C/C++ programming.
In some cases, Python may be useful.

- Regular attendance of all sessions of the seminar according to the list of attendees.
- Successful seminar talk on the assigned topic. The presentation can be up to 45 minutes. The presentation is followed by a discussion on the presented topic with the presenter which can be up to 15 minutes. Presentation language is English; Discussion is in English or German (depending on the audience).
- A presentation should include prototypes and/or brief benchmarks of the selected tools or libraries.
- A short written report, lecture notes or "how-to" (depending on the topic) of the presented topic.
- Active and meaningful contributions to all discussions of the seminar.

Places: 10

Aktuelle Probleme von Recht und Digitalisierung des Straßenverkehrs by Dr. Jochen Krüger, Prof. Christoph Sorge, Stephanie Vogelgesang

Die Digitalisierung des Straßenverkehrs schreitet unaufhörlich voran. Dadurch entstehen neue technische, aber auch neue rechtliche Fragestellungen. Aktuelle Beispiele dafür sind die Auswertung von Fahrzeugdaten oder der Bereich des automatisierten Fahrens.

Es handelt sich um ein interdisziplinäres Seminar für Juristen und Informatiker. Aufgabe der Teilnehmer ist es, eine Seminararbeit zu einem ausgewählten Thema zu verfassen und diese in einem Vortrag vorzustellen. Der Vortrag fließt neben der Seminararbeit in die Gesamtnote ein.

Weitere Informationen sowie Themenvorschläge finden Sie auf der Webseite des Lehrstuhls unter legalinf.de

Requirements: Es wird erwartet, dass Teilnehmer in der Lage sind, Vorträgen in deutscher Sprache zu folgen.

Places: 7

Comparing Fuzzing Techniques by Andreas Zeller, Rahul Gopinath

In this seminar, we evaluate automated test generation techniques (fuzzers) for their effectiveness and efficiency. We discuss and design evaluation criteria and apply them on a number of techniques from the book "Generating Software Tests" (https://www.fuzzingbook.org). Apart from reporting and presenting your results in the seminar, your evaluation results will be included in the book. If the results or techniques are novel (many of them are), we will also strive to publish them as a scientific paper, with you as co-author.

Group phase:
* Introduction to empirical techniques (all)
* Designing benchmarks (all)
* Defining common evaluation questions (all)
* Presenting empirical results (all)

Individual phase:
* Selecting an individual fuzzing technique for evaluation
* Defining individual evaluation questions and parameters
* Conducting the evaluation
* Producing an evaluation report
* Presenting and discussing evaluation results

The fuzzers are implemented in Python and come as Jupyter Notebooks. You will also use Jupyter Notebooks for conducting the evaluation and presenting the final results, both as report and as presentation.

Requirements: You should have passed the lecture "Generating Software Tests" or have sufficient knowledge from the course.

Knowledge of Python and experience with Jupyter Notebooks is helpful, but can be acquired during the seminar.

Places: 10

Computer Vision and Machine Learning for Computer Graphics by Christian Theobalt, Mohamed Elgharib, Vladislav Golyanik

Computer Vision strives to develop algorithms for understanding, interpreting and reconstructing information about real-world scenes from image and video data. Computer Graphics focuses on image synthesis – algorithms to build and edit static and dynamic virtual worlds and to display them in photorealistic or stylised ways. In recent years, these fields have converged more and more since both disciplines create and exploit models describing the visual appearance of objects and scenes. The objective of the seminar "Computer Vision and Machine Learning for Computer Graphics" is to cover various topics related to applications of machine learning in computer graphics as well as methods on the intersection of computer vision and computer graphics. We will focus on the recent results published at top-tier conferences and journals. The examples of the covered research fields are: monocular face reconstruction and reenactment, hand tracking, video motion magnification, monocular non-rigid 3D reconstruction, human pose estimation, deep convolutional networks, and several others.

Visit http://gvv.mpi-inf.mpg.de/teaching/gvv_seminar_2019/ for more details!

Attention: the already registered students (before the SIC portal has started to operate) are already accepted for the seminar.

Requirements: Knowledge of at least one of the following topics is essential: optimisation methods, image processing, 3D geometry, computer vision or computer graphics. Expected: knowledge in linear algebra and calculus.

Places: 8

Hybrid Systems (Block Seminar) by Prof. Bernd Finkbeiner, Maximilian Schwenger, Jesko Hecking-Harbusch, Noemi Passing, Malte Schledjewski

The interaction of a computer system with physical processes is a clash of two worlds: a computer works in a clocked, discrete world with clear semantics. When the processor's edge rises, the internal state changes, when it falls, the world stands still. The real, physical world, however, is chaotic: nature does not wait for a clock tick to apply changes. Yet, the interaction of machines with the real world is essential. Machines control power plants, cars, satellites and factories. Such systems are called "Hybrid Systems", a mixture of continuous, physical processes and discrete, logical steps.

In this seminar we will cover an interesting range of topics concerning hybrid systems such as analysis methods, synthesis and runtime monitoring.

The seminar spans over four weeks, starting on August 12, and ending September 6. The Kick-off meeting will take place on August 12, 10am.

You can find more information here: https://courses.react.uni-saarland.de/hs19/

Requirements: The seminar is for advanced bachelor or master students. Basic mathematical knowledge (e.g. Mathematik für Informatiker I-III) is required.

Places: 12

Legal Tech by Stefan Hessel, Dr. Jochen Krüger, Prof. Christoph Sorge, Stephanie Vogelgesang

Legal Tech bezeichnet den Einsatz von Informationstechnik im juristischen Bereich. Insbesondere geht es um die Nutzung neuer Technologien sowie künstlicher Intelligenz durch die Rechtsanwaltschaft und neuerdings auch durch die Gerichte.

Weitere Informationen sowie Themenvorschläge finden Sie auf der Webseite des Lehrstuhls unter legalinf.de

Requirements: Es wird erwartet, dass Teilnehmer in der Lage sind, Vorträgen in deutscher Sprache zu folgen.

Places: 7

Optimization for Machine Learning by Peter Ochs

In this seminar, we discuss recent scientific papers in the field of mathematical optimization with a focus on problems from Machine Learning and related fields such as Computer Vision, Image Processing, Compressed Sensing or Statistics. The topics may include all aspects of optimization: algorithms, analysis, applications.

The topics are listed here: http://www.mop.uni-saarland.de/teaching/SOPTM19/index.shtml#material

Format: Each student is assigned to three papers (assigned by us): one student gives the presentation and writes a short summary on the paper, and the two other students take the role of opponents, i.e., they prepare questions and lead the discussion after the presentation.

Requirements: The seminar is for advanced bachelor or master students in Mathematics or Computer Science. Requirements are basic mathematical knowledge (e.g. Mathematik für Informatiker I-III) and a course on optimization, for example, "Continuous Optimization" or "Convex Optimization".

Places: 1

Programming Languages and Compilers for Machine Learning by Sebastian Hack

Machine learning brings interesting new challenges for programming languages research in terms of making machine learning applications execute more efficiently but also in new different programming paradigms that are specifically tailored to machine learning tasks. In this seminar, we are reading recent papers on programming systems for machine learning, mostly deep learning. The papers we read deal with program representations (IRs) and program analyses for neural nets, domain-specific languages for tensor calculus, and differentiable programming.

Requirements: It is beneficial to have some background in compilers (i.e. by having passed the compilers core course). You don't need to have a background in machine learning. We will cover the necessary foundations in the seminar.

Places: 12

Secure Compilation (BLOCK) by Marco Patrignani

Please check the seminar page below for all seminar-related information:

https://cms.cispa.saarland/scs19/

Requirements: Basic knowledge of programming language semantics is a bonus, though lectures cover that as well.

Places: 19

Selected Topics in Mobile Security by Sven Bugiel

In this seminar, we will discuss current results and new problems in the mobile security domain based on relevant scientific papers. The focus of the selected papers lies on Android, given its high popularity among researchers. The topics include usability aspects of Android's permission system and security-relevant APIs, security extensions at different levels of Android's software stack, app analysis, and newly identified attack vectors.

Requirements: There are no formal pre-requisites beyond a basic knowledge of how computer systems work internally. Having worked with a smartphone as user or app developer, or participation in the Advanced Lecture Mobile Security are beneficial.

Places: 12

Semantic Deep Learning by Matthias Klusch, Igor Vozniak, Farzad Nozarian

This seminar is concerned with selected methods and systems of semantic deep learning. Research on semantic deep learning mainly focuses on the task-oriented combination of deep learning with symbolic knowledge representation and reasoning in general, and semantic web technologies in particular. In addition, the rationalization and semantic explanation of deep learning models and results to the human user is of interest (explainable AI). In this seminar, we will take a closer look at advanced methods, systems and tools for semantic deep learning, and discuss their strengths and weaknesses.

The seminar type is classic in the sense that registered participants will present the assigned topics, and discuss the strength and weaknesses of presented approaches. In addition, there will be two dedicated opponents for each presentation of an assigned topic. Short individual tutoring sessions for presenters of assigned topics by seminar tutors are offered on request.

The seminar is held weekly on Thursday from 4pm (16:00) to 6pm at DFKI in room “Turing 1” (NB R +2.30). The location can be accessed via the entrance of the DFKI main building D3.2. You may ask at the reception in the lobby of the building for further directions.

For more information on (pre-)registration, topics, schedule and tutors of the seminar, please visit the seminar page:

http://www.dfki.de/~klusch/SemDeep-seminar-ss19/

Requirements: This seminar aims primarily at advanced master students in Computer Science who preferably hold a B.Sc. degree in this or related field. Good knowledge in AI, in particular not only in deep learning but also basics of knowledge representation and reasoning is required. Selected background references for reading, if required, are given on the topic page.

Requirements for successfull participation:

* Confirmed pre-registration to the seminar by email from the seminar leader (until April 11, 2019 latest), and then registration to the seminar in HISPOS (until May 6, 2019 latest).

* Regular attendance of all sessions of the seminar according to the list of attendees

* Successful seminar talk on the assigned topic. The presentation can be up to 45 minutes. The presentation is followed by a discussion on the presented topic with the presenter (defense) which can be up to 45 minutes. This discussion with the audience is particularly driven by two dedicated opponents. Presentation language is English or German (depending on audience); the presentation slides have to be in English.

* Active and meaningful contributions to all discussions of the seminar. That includes the opposition task mentioned above. Opponents are supposed to carefully prepare for their in-depth discussion with the respective presenter about the presented methods and applications.

Weighted evaluation of participation (in %):
Presentation and defense 75, Discussion and opposition 25

Strafrechtliche Betrachtung von Cyberangriffen by Stefan Hessel, Dr. Jochen Krüger, Prof. Christoph Sorge, Stephanie Vogelgesang

Die allgemeine Digitalisierung der Gesellschaft zwingt dazu, sich auch mit technischen und strafrechtlichen Aspekten von Cyberangriffen zu befassen. Vor diesem Hintergrund beschäftigt sich das Seminar mit ausgewählten Formen der digitalen Kriminalität.

Weitere Informationen sowie Themenvorschläge finden Sie auf der Webseite des Lehrstuhls unter legalinf.de

Requirements: Es wird erwartet, dass Teilnehmer in der Lage sind, Vorträgen in deutscher Sprache zu folgen.

Places: 7