Seminar

Inpainting-based Image Compression

Winter Term 2025 / 2026

(Main) Seminar: Inpainting-based Image Compression

Prof. Joachim Weickert

Winter Term 2025 / 2026

(Main) Seminar (2 h)

Notice for bachelor/master students of mathematics: This is a »Hauptseminar« in the sense of these study programs.

Image reconstruction from a sparse subset of its data (located in the left image). Author: P. Peter

Important Dates – Description – Registration – Requirements – Introductory Meeting – Overview of Topics

Introductory meeting (mandatory):
The introductory meeting took place on Wednesday, November 5, 2025, 16:15 via Zoom. The link has been sent to all seminar participants. In this meeting, the papers have been assigned to the students. Attendance was mandatory for all participants.

Registration deadline in the LSF system:
The deadline is November 27.

Regular meetings during the winter term 2025 / 2026:
Wednesday 16:15-18:00, online via our Zoom link.

Overview talk by Joachim Weickert:
November 26, 2025

First student presentations:
December 3, 2025.

Last student presentations:
January 21, 2026.

Write-up submission deadline:
March 1, 2026, 23:59. Please e-mail your write-up to Joachim Weickert.

Contents: Classical lossy image compression methods such as JPEG rely on the idea to simplify information in a frequency or wavelet domain. These established methods can be challenged by an alternative class of algorithms: Inpainting-baed methods work entirely in the spatial domain. They store only a small but carefully optimised subset of the image data. In the decoding phase, the missing image data are filled-in with so-called inpainting approaches. While this idea sounds surprisingly simple, it gives rise to many interesting questions and challeging problems. In this seminar we will cover the various aspects, from the basics to state-of-the art approaches.

Prerequisites: The seminar is for advanced bachelor or master students of a mathematical or computer science study program. Basic mathematical knowledge (e.g. Mathematik für Informatiker I-III) is required. It is recommended that applicants have passed the core class on "Image Processing and Computer Vision". They should mention in their application if this is the case.

Language: All papers are written in English, and English is the language of presentation.

Applications to this seminar were possible via the SIC Seminar Assignment Website until October 14, 2025, 23:55. The seminar is fully booked now.

To administrate their grades, admitted students must register in the LSF system by November 27.

Regular attendance: You must attend all seminar meetings, except for provable important reasons (medical certificate).

Talk: Talk duration is 30 min, plus 15 min for discussion. Please do not deviate from this time schedule.
Your presentation must be given in English. Your slides and your write-up, too, should be in English.

Write-up: The write-up has to submitted until Sunday, March 1, 2026, 23:59 via e-mail to Joachim Weickert. It should summarise your talk on 5 pages. The file format for electronic submissions is PDF – text processor files (like .doc) are not acceptable. Participants who do not submit a writeup cannot obtain the certificate for the seminar.

Plagiarism: Adhere to the standards of scientific referencing and avoid plagiarism: Quotations and copied material (such as images) must be clearly marked as such, and a bibliography is required. Otherwise the seminar counts as failed.

Mandatory consultation: Talk preparation has to be presented to your seminar supervisor in the week before the talk is given. It is your responsibility to make your appointment two weeks ahead of your talk.

No-shows: No-shows are unfair to your fellow students: Some talks are based on previous talks, and your seminar place might have prevented the participation of another student. Thus, in case you do not appear to your scheduled talk (except for reasons beyond your control), we reserve the right to exclude you from future seminars of our group.

Active participation in discussions: The discussions after the presentations are a vital part of this seminar. This means that the audience (i.e. all paricipants) poses questions and tries to find positive and negative aspects of the proposed ideas and the presentation itself. This activity is part of your final grade.

Being in time: You are expected to participate from the beginning to the end.

Here are the (slightly updated) slides from the introductory meeting. They contain important information for preparing a good talk.

We will discuss the papers below. If your registration was successful, the password has been sent to you before the introductory meeting. This enabled you to inspect the papers and identify your favourites. Since this area has been poineered by our group and we have received an ERC Advanced Grant to support scientific progress in this field, we focus on articles that describe our contributions. The slides of the talks will be added when they are available.

Contour-based Approaches

Data Optimisation

Basic Codecs and Efficient Algorithms

Applications Beyond Still Image Coding

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

Imprint - Data protection