Modern Methods in Image Analysis

Winter Term 2008/2009

Modern Methods in Image Analysis

Dr. Andrés Bruhn, Prof. Joachim Weickert

(bruhn@mia.uni-saarland.de)

Winter Term 2008/2009

Seminar (2h)

Important Dates – Description – Administrative details – Requirements – Overview of Topics

First meeting: Friday, July 11, 2008, 14:15h, Bld. E1.1, room 306

Sign-up deadline: Thursday, July 10, 2008.
No free capacity is left in this seminar.

Regular meetings during winter term 2008/2009:
Tuesdays, 16:15h , Bld. E1.1, room 306

Date of first regular meeting: Tuesday, November 25.

Please direct all organisatorial requests to Dr. Andrés Bruhn.

Contents: There has been enormous progress in image processing and computer vision in recent years. Both the quality of the models and the efficiency of the algorithms have improved significantly. In this seminar the participants learn more about such modern techniques that make use of sophisticated strategies. The topics include but are not limited to

• optic flow

• stereo reconstruction

• shape from shading

• dynamic textures

• image denoising

Prerequisites: The seminar is designed for graduate students (i.e. advanced Bachelor, Master, and post-Vordiplom) of visual computing, computer science or mathematics. Specific knowledge in image analysis or computer visision is recommended.

Language: Understanding scientific text in English is indispensable. Talks and write-up can be in English or German.

Sign-up: If you are interested to participate, please sign up by e-mailing to bruhn@mia.uni-saarland.de not later than Thursday, July 10, 2008. Please include your name, student ID (Matrikelnummer), date of birth, study program, and valid e-mail address.
Since the number of talks is limited, we ask for your understanding that participants are considered strictly in the order of incoming e-mails – no exceptions.

Sessions: First meeting takes place Friday, July 11, 2008, 14:15h, Bld. E1.1, room 306. Regular sessions will be weekly during the winter term. Exact schedule to be determined at first meeting.

Regular attendance: It is expected that you attend all seminar meetings, except for provable important reasons (we check).

Talk: Talk duration is 30 min, plus 15 min for discussion. Please do not deviate significantly from this time schedule.
You may give a presentation using a data projector (»beamer«), overhead projector or blackboard, or mix these media appropriately. English is preferred as language of presentation; if you should decide to talk in German anyway, you must use slides in English.

Write-up: The write-up has to be handed in till the end of the semester A concrete data will be announced at the first regular seminar meeting. The write-up should summarise your talk, so some 5+ pages per speaker will be adequate in most cases. Electronic submission is preferred. File format for electronic submissions is PDF – text processor files (like .doc) are not acceptable. We recommend using LaTeX. Adhere to the standards of scientific referencing: Quotations and copied material (such as images) must be clearly marked as such, and a bibliography is required.

Mandatory consultation: Talk preparation has to be presented to your seminar supervisor not later than in the week before the talk is given. It's your responsibility to approach us timely and make your appointment!

Please note that the presentation dates are preliminary and subject to change.

No.	Topic	Date
1	A1 - Optic Flow and Parameterized Motion A3 - Optic Flow and Segmentation	Nov 25
2	A4 - Optic Flow and Real-Time Numerics D1 - Optic Flow and Dynamic Textures	Dez 2
3	B1 - Registration of Tensor Fields B2 - Registration with Joint Intensity Priors	Dez 9
4	C1 - Stereo and Occlusion Handling C2 - Stereo and Fundamental Matrix Estimation	Dez 16
5	C3 - Stereo and Scene Flow C4 - Stereo and Range Images	Jan 13
6	F1 - Shape from Shading G2 - Denoising with Bilateral Filter	Jan 20
7	G1 - Denoising with Non-Local Means G3 - Denoising with Unified Smoothing Approach	Jan 27

A1	T. Nir, A. Bruckstein, R. Kimmel Over-Parameterized Variational Optical Flow International Journal of Computer Vision (IJCV), 2008
A3	T. Brox, A. Bruhn, J. Weickert Variational Motion Segmentation with Level Sets European Conference on Computer Vision (ECCV), 2006
A4	A. Bruhn, J. Weickert Towards Ultimate Motion Estimation: Combining Highest Accuracy with Real-Time Performance International Conference on Computer Vision (ICCV), 2005
D1	T. Amiaz, S. Fazekas, D. Chetverikov, N. Kiryati Detecting Regions of Dynamic Texture International Conference on Scale Space and Variational Methods in Computer Vision (SSVM), 2007
B1	S. Barbieri, M. Welk, J. Weickert Variational Registration of Tensor-Valued Images. Workshop Tensors in Image Processing and Computer Vision IEEE Conference on Computer Vision and Patteren Recognition, 2008
B2	D. Cremers, C. Guetter, C. Xu Nonparametric Priors on the Space of Joint Intensity Distributions for Non-Rigid Multi-Model Image Registration. IEEE Conference on Computer Vision and Patteren Recognition (CVPR), 2006
C1	R. Ben-Ari, N. Sochen Variational Stereo Vision with Sharp Discontinuities and Occlusion Handling. International Conference on Computer Vision (ICCV), 2007
C2	L. Valgaerts, A. Bruhn, J. Weickert A Variational Approach for the Joint Recovery of the Optical Flow and the Fundamental Matrix. Annual German Pattern Recognition Symposium (DAGM), 2008
C3	F. Huguet, F. Devernay A Variational Method for Scene Flow Estimation from Stereo Sequences. International Conference on Computer Vision (ICCV), 2007
C4	C. Zach, T. Pock, H. Bischof A Globally Optimal Algorithm for Robust TV-L1 Range Image Integration. International Conference on Computer Vision (ICCV), 2007.
F1	O. Vogel, A. Bruhn, J. Weickert, S. Didas Direct Shape-from-Shading with Adaptive Higher Order Regularisation. International Conference on Scale Space and Variational Methods in Computer Vision (SSVM), 2007
G2	C. Tomasi, R. Manduchi Bilateral Filtering for Gray and Color Images. International Conference on Computer Vision (ICCV), 1998.
G1	A. Buades, B. Coll, J.-M. Morel A Non-Local Algorithm for Image Denoising. IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2005.
G3	P. Mrázek, J. Weickert, A. Bruhn On Robust Estimation and Smoothing with Spatial and Tonal Kernels. Geometric Properties from Incomplete Data, 2006.

Andrés Bruhn / July 4, 2008

MIA Group
©2001-2023
The author is not
responsible for
the content of
external pages.

Imprint - Data protection