• Home
• Research
• Publications
• Software
• Teaching 
• People
• Staff Only
• Events
• Open Positions
• Links
• Contact

Fred A. Hamprecht

Fred A. Hamprecht Robert-Bosch endowed Full Professor for Multidimensional Image Processing Affiliate Professor of Pathology, Children's Hospital, Boston

Office G.2.01
Heidelberg Collaboratory for Image Processing (HCI)
Interdisciplinary Center for Scientific Computing (IWR)
Universität Heidelberg
Speyerer Straße 6
D-69115 Heidelberg

Tel: ++49-6221-54 88 00
Sec: ++49-6221-54 88 75
Fax: ++49-6221-54 52 76
fred.hamprecht@iwr.uni-heidelberg.de

next office hours:
Tue 19.01. 14:00
Mon 25.01. 14:00
Tue 02.02. 14:00

Teaching

• Main course winter semester 2009: Pattern Recognition / Mustererkennung
• Main course summer semester 2009: Imaging physics / Physik der Bildgebung
• Main course winter semester 2008: Image Processing / Bildverarbeitung

Research / Scientific Interests

Methodological:

• Variants of supervised learning
  • weakly supervised learning
  • semi-supervised learning
  • active learning
• Inference using spatial context

Main applications:

• Industrial quality control
• Data analysis in the life sciences: spectroscopic images, high-throughput microscopy
• Proteomics: data mining of mass spectrometric data

Collaborations

I am greatly indebted to my colleagues with their different backgrounds: if work is fun and fruitful, it is thanks to them! Beyond my group and the HCI, the most intensive current cooperations are with (ordered by first letter)

• Hanno and Judith Steen, Children's Hospital / Harvard Medical School, Boston
• Margareta Mueller, dkfz, Heidelberg
• Matthias Mayer, ZMBH, Heidelberg
• Oliver Nix, Peter Bachert, Wolfgang Schlegel, dkfz, Heidelberg
• Ron Heeren, AMOLF, Amsterdam
• Winfried Denk and Moritz Helmstaedter, MPI for Medical Research, Heidelberg

Some Lecture Notes

Nowadays I prefer to deliver lectures on the blackboard. This collection of some older slides generally does not reflect the contents of the current lectures.

• Introduction to Statistics
• Signal processing
  • (pdf slides 2 up) (pdf slides 4 up) Introduction to signal processing, Fourier transform
  • (pdf slides 2 up) (pdf slides 4 up) Time-invariant systems, z-transform, filters
  • (pdf slides 2 up) (pdf slides 4 up) B-spline signal processing
  • (pdf slides 2 up) (pdf slides 4 up) Parametric spectral density estimation
  • (pdf slides 2 up) (pdf slides 4 up) Nonparametric spectral density estimation
  • (pdf slides 4 up) Hilbert transform (courtesy of M. Hissmann)
  • (pdf slides 2 up) (pdf slides 4 up) Time-frequency decompositions
  • (pdf slides 2 up) (pdf slides 4 up) Continuous wavelet transform
  • (pdf slides 2 up) (pdf slides 4 up) Multiresolution analysis
  • (pdf slides 2 up) (pdf slides 4 up) Discrete wavelet implementation and applications
  • (pdf slides 2 up) (pdf slides 4 up) Hidden Markov Models
  • (pdf slides 2 up) (pdf slides 4 up) Optimal discrete filters
  • (pdf slides 2 up) (pdf slides 4 up) Adaptive filters
  • (pdf slides 2 up) (pdf slides 4 up) Sampling and interpolation

• Image Processing
• (pdf slides 2 up) (pdf slides 4 up) Introduction to image processing; color spaces; image formats: TIFF, GIF, PNG
• (pdf slides 2 up) (pdf slides 4 up) Linear operators; discrete Fourier transform; discrete cosine transform; JPEG compression
• (pdf slides 2 up) (pdf slides 4 up) Filters and their optimization
• (pdf slides 2 up) (pdf slides 4 up) Linear isotropic diffusion filtering; Gaussian and Laplacian pyramid and their efficient implementation
• (pdf slides 2 up) (pdf slides 4 up) Anisotropic (Perona-Malik) diffusion filtering; structure tensor and coherence-enhancing diffusion
• (pdf slides 2 up) (pdf slides 4 up) Bayesian image analysis; Markov and Gibbs random fields; Markov Chain Monte Carlo methods
• (pdf slides 2 up) (pdf slides 4 up) Gibbs sampler, Metropolis sampler, minimization heuristics: iterated conditional modes (ICM), relaxation labeling (RL), highest confidence first (HCF), graduated non-convexity (gnc)
• (pdf slides 2 up) (pdf slides 4 up) Applications of Bayesian image analysis: smoothing, restoration and deconvolution; object matching; spin glass Markov random fields in object recognition
• (pdf slides 2 up) (pdf slides 4 up) Texture description and recognition: filter banks and textons, the role of vector quantization, joint distribution of intensities, Haralick and Unser parameters
• Pattern recognition
  • (pdf slides 2 up) (pdf slides 4 up) Introduction to pattern recognition; linear and quadratic discriminant analysis (LDA, QDA)
  • (pdf slides 2 up) (pdf slides 4 up) Statistical decision theory; nearest neighbor methods
  • (pdf slides 2 up) (pdf slides 4 up) Model selection and assessment: bias-variance trade-off, resampling methods: cross-validation, bootstrap
  • (pdf slides 2 up) (pdf slides 4 up) Perceptron and multilayer-perceptron (feed-forward neural networks)
  • (pdf slides 2 up) (pdf slides 4 up) Support Vector Machines
  • Ordinary least squares, Regularized Regression Methods
  • Total least squares / errors-in-variables
  • Outlier diagnostics and robust regression
  • (pdf slides 2 up) (pdf slides 4 up) Kernel density estimation
  • (pdf slides 2 up) (pdf slides 4 up) Cluster analysis
  • Multidimensional Scaling
  • Nonlinear dimension reduction
  • (pdf slides 2 up) (pdf slides 4 up) Local linear embedding; bilinear decomposition

Last update: 08.01.2010, 18:25

Webmaster: Ole Hansen