Riemannian Geometric Statistics in Medical Image Analysis (Seminar)

If you are interested, please write a short e-mail to jonathan.schwarz@iwr.uni-heidelberg.de.

Please write a mail to jonathan.schwarz@iwr.uni-heidelberg.de until October 31st, 2020 if you want to participate.

The Proseminar will take place every Friday from 4-6 pm starting in November.

If you want to participate in the seminar, you have to:

• give a talk (using the board or slides …)
• hand in a written summary latest a week after your talk
• attend the other talks

If there are questions upcoming during the preparation of the talk, please don't hesitate to ask.

You have to give a talk on your topic

• The talk will be done in presents or via zoom (depending on the Covid situation)
• You are free to choose any format for your presentation (slides or writing some notes during the presentation)
• The talk should last between 20 and 30 minutes.
• There will be a discussion session after your talk of around 5 to 10 minutes

Send the slides as PDF to jonathan.schwarz@iwr.uni-heidelberg.de the latest the day before your talk.

You have to hand in a written summary of your topic the latest 7 days after your talk.

• The summary should be created with LaTeX and should be handed in as a Pdf file.
• I recommend the LNCS LaTeX Template, but you don't have to use it.
• The summary should be between 2 and 4 pages long

Send the summary as PDF to jonathan.schwarz@iwr.uni-heidelberg.de.

• Submission slides: at least one day in advance of your presentation

• Submission summary: one week after of your presentation

Part of the seminar will be based on the book Riemannian Geometric Statistics in Medical Image Analysis by Xavier Pennec, Stefan Sommer and Tom Fletcher. You will get access to the book. Possible topics are: * Manifold-valued image processing with SPD matrices * Riemannian geometry on shapes and diffeomorphisms * Beyond Riemannian geometry * Low-dimensional shape analysis in the space of diffeomorphisms

• SymNMF: nonnegative low-rank approximation of a similarity matrix for graph clustering by Kuang, Yun and Park, Journal of Global Optimization, 2015, pdf
• k-MLE: A fast algorithm for learning statistical mixture models by Frank Nielsen, arXiv preprint, 2012, pdf
• Wasserstein Dictionary Learning: Optimal Transport-Based Unsupervised Nonlinear Dictionary Learning by Schmitz et al., SIAM Journal on Imaging Sciences, 2018, pdf
• Ising and Potts models on the hypercubic lattice by Duminil-Copin H., arXiv preprint arXiv:1707.00520, 2017 link (only one of the chapters 1, 2, 4, 5.1, 6.1)
• Information geometry connecting Wasserstein distance and Kullback–Leibler divergence via the entropy-relaxed transportation problem by Shun-ichi Amari et. al., Information Geometry, Springer, 2018, pdf
• Escort Evolutionary Game Theory by Marc Harper, arXiv, 2012, pdf