The digital processing of visual information has become a core topic in modern
CS and IT. Visual Computing (VisComp) builds upon foundations from Computer
Science and applied Mathematics and has a wide range of applications.
Methodologically, VisComp is routed in computer graphics, algorithmic geometry,
image processing and computer vision as well as machine learning. Given the
importance and international standing of VisComp we believe that our students
should have an opportunity to acquire fundamental knowledge via a new major.


The proposed major will acquaint students with core knowledge in visual
information processing and learning. We will cover the most important
algorithmic and systems oriented foundations and form a basis for more
specialized courses in computer graphics, digital image processing and machine
learning. Topics will include:

Fundamentals 

a) Digital signal representation, sampling theorem
b) Grey value, color and
vector quantization c) Color spaces d) Transformations and projections e)
Projective geometry, homogeneous coordinates f) Camera models g) Calibration h)
Lighting and shading i) Texture mapping j) Scan conversion k) Graphics APIs and
graphics hardware


Computer Vision 

a) Digital image processing: image restoration
b) Image data formats:

LaplaceGauss pyramid

wavelets

nonlinear diffusion
c) Edge detection
d) Image segmentation
e) Optical flow
f) Stereo vision
g) Shape from X
h) 2d & 3d Form models
i) Statistic image models, MRFs
j) Object recognition
k) Texture


Advanced Rendering 

a) Global illumination

ray tracing

radiosity

stochastic methods
b) Rendering of point models, surfels
c) Imagebased rendering
d) Levelofdetail rendering
e) Hierarchical texturing


3D Modeling and Geometry Processing 

a) Parametric curves, Bézier curves, Bsplines, NURBS
b) Tensor product surfaces, triangle meshes
c) Mesh fairing
d) Subdivision methods
e) Point models
f) Convex hulls, Delaunay triangulations
g) Spatial search structures
h) Physicsbased modeling


Physicallybased Simulation 

a) Massspring models
b) Particle systems
c) Rigid body dynamics
d) Finite element method
e) Collision detection
f) Fluid simulation
g) Applied PDEs


Scientific Visualization 

a) Contouring and isosurfaces
b) Direct volume rendering
c) Streamline integration
d) Texture advection
e) Flow topology
f) Feature extraction
g) Tensor field visualization
h) Information visualization


Supervised Learning 

a) Bayesian decision theory
b) Classifier design:
 knn classifiers
 perceptron learning
 support vector machines
 LDA
 neural networks
 boosting
c) Graphical models:
 belief propagation
 variational methods
 hidden Markov models
d) Regression
e) Statistical learning theory (COLT)


Unsupervised Learning 

a) Custering and mixture models
b) Markov random fields.
c) reinforcement learning
d) maximum entropy inference
e) Neuroinformatics:
 organization principles of neural systems
 dynamics of neurons and neural assemblies
 synaptic plasticity




