Teaching

Summer Term 2018

Lecture: Grundlegende Programmiertechniken (German)

V2+U2
Lecture Time & Place: Tue. 10:15-11:45 c.t., LB 107
Instructor: Prof. Dr. J. Krüger
Exercise Time & Place: see LSF
Exercise Supervisor: Contact at gpt@uni-due.de (A. Krekhov)
Course Language: German
Audience: Bachelor

Zusammenfassung: Anhand der Programmiersprachen Python 3 und Java werden grundlegende Programmiertechniken besprochen. Inhalte unter anderem - Einführung und grundlegende Struktur von Programmen - Lexikalische Elemente, Datentypen und Variablen, Anweisungen und Kontrollstrukturen - Elementare Datenstrukturen: Mengen, Felder, Listen - Eingabe und Ausgabe mittels Streams - Funktionen und Module - Grundlegende Algorithmen: Suchen und Sortieren - Eingabe/Ausgabe Listen-Abstraktion. Unterschiede Java / Python.

Lecture: Scientific Visualization

V2+U2
Lecture Time & Place: Mon. 10:00-12:00 c.t. LE 120
Instructor: Prof. Dr. J. Krüger
Exercise Time & Place: Mon. 12:00-14:00 c.t. LE 120
Exercise Supervisor: K.-M. Schorer
Course Language: English
Audience: Master

Abstract: The focus of this introductory course is on discussing efficient techniques to visually represent large-scale data sets from simulation and measurement. Starting with a brief introduction on the data generation processes the visualization pipeline, data structures, mapping techniques and special rendering techniques for scientific data will be discussed. Various examples will be given to outline the benefits of visualization techniques in practical applications. A particular focus is put on interactive methods using GPU-based techniques. Topics include:

History of Visualization Methods, from ancient wall paintings to GPU accelerated rendering
Data Sources

Simulation: basics of finite element discretization to GPU-based methods
Measurements: physical details of computed tomography to accelerated reconstruction

Data Representation: interpolation techniques, from data to meshes
Filtering: various filtering methods and their hardware acceleration
Scalar Volumes

the basics of direct volume rendering
GPU accelerated volume rendering I: slicing and ray-casting
GPU accelerated volume rendering II: ray guided volume rendering and out-of-core techniques
indirect volume rendering
unstructured data

Flow Visualization

glyph based methods
geometric methods
GPU-based particle tracing
texture-based methods I: LIC
texture-based methods II: Explicit Frequency Control
feature based methods

Virtual Reality

Lecture: Advanced Image Synthesis

V3+U1
Lecture Time & Place:
Tue. 14:00-16:00 c.t. LE 120, Thu. 9:00-10:00 c.t. LE 105
Instructor: Prof. Dr. J. Krüger
Exercise Time & Place: Thu. 8:00-9:00 c.t. LE 105
Exercise Supervisor: A. Waschk
Course Language: English
Audience: Master

Abstract: This lecture deals with the fundamentals of photorealistic and interactive image synthesis. In particular, we discuss techniques to achieve interactive frame rates for the realistic rendering of complex models and scenes using GPU accelerated techniques. Selected graphics algorithms and their efficient implementation exploiting state-of-the-art graphics hardware through graphics APIs will be demonstrated. We focus on dedicated parts of the graphics pipeline and we review the functionality provided by consumer class graphics accelerators including programmable vertex-, geometry- and fragment shaders. In addition, we discuss the governing equations in physics based light transport and we demonstrate effective solution methods for the simulation of global illumination. Topics include:

Interactive image synthesis

Graphics APIs & hardware, OpenGL, DirectX, Vulkan and CUDA
Fixed function & programmable graphics pipeline
Introduction to General-purpose computation on graphics hardware (GPGPU)
Lighting

Local lighting models
GPU implementation

Environment Mapping
GPU Shadowing techniques

Projective Shadows
Shadow Volumes
Shadow Maps

Transparent Objects:

(depth) sorting with implementations on graphics hardware
Order independent transparency & depth peeling on GPUs

Physics based rendering

Radiometric quantities, rendering equation
Raytracing / path tracing
Radiosity
Irradiance volumes
Precomputed radiance transfer (PRT)
Ambient Occlusion with focus on GPU-based interactive methods

Special topics:

GPU friendly High quality terrain rendering & synthesis
Rendering of fur and hair
Scenegraph APIs
Acceleration structures and culling techniques
Human visual perception and high dynamic range imaging (HDRI)
REYES
Image compression techniques

Lab Course (Praxisprojekt): Dear Data 1

Kick-off & Registration Meeting: April 10, 13:00 s.t., LE 335
Time & Place: tba
Instructor: A. Krekhov & Prof. Dr. J. Krüger
Course Language: German
Audience: Bachelor

Abstract: In this lab course the participants will learn to gather data and design scientific visualisations using this data. We will use an approach similar to the award winning Dear Data project by Giorgia Lupi and Stefanie Posavec. Given a specific topic each week, participants will have to track their data and custom design a visual representation of their data and present this representation in front of the group at the end of the week.

Lab Course (Praxisprojekt): Dear Data 2

Kick-off & Registration Meeting: April 10, 12:00 s.t., LE 335
Time & Place: tba
Instructor: M. Michalski & Prof. Dr. J. Krüger
Course Language: German
Audience: Bachelor

Lab Course (Forschungsprojekt): Smart Running

Kick-off & Registration Meeting: April 12, 10:00 s.t., LE 335
Time & Place: Thu. 10:00-12:00 s.t., LE 335
Instructor: A. Schiewe & Prof. Dr. J. Krüger
Course Language: English or German
Audience: Master

Abstract: More and more runners record their runs with fitness trackers, running watches, or smartwatches. In addition to these watch-based sensing devices worn on the wrist there is an increase in other wearable devices and sensors (STRYD, RunScribe, Arion, SHFT IQ) that attempt to support the runner in improving their performance or running technique. In this project we will evaluate fundamental concepts and develop novel interfaces for runners using commercial or self-made sensor prototypes. The implementation of the project can range from the evaluation of interface sketches which require no programming skills to the implementation of real functioning wearable prototypes depending on the skill set and motivation of the participants.

Seminar: How to Deliver an Engaging Speech - Eine Einführung in elementare Vorbereitungs- und Vortragstechniken

Instructors: Prof. Dr. J. Krüger & team
Kick-off & Registration Meeting: Thu. April 12th, 12:00-14:00 c.t., LF 125
Time & Place: Thu. 12:00-14:00 c.t., LF 125
Course Langugage: English or German
Audience: Bachelor & Master

Abstract: In addition to the scientific content in this seminar we will put special emphasis on the presentation. To improve the presentation skills of the participants we will start the seminar with simpe topics, record short talks from each participant, and analyze the presentation. Students then give a seminar talk on their chosen topic.