Graphics, Vision and Video - Interdisciplinary Topics in Visual Computing

Seminar – Summer Semester 2012

Organizer: Christian Theobalt, Co-organizers: Levi Valgaerts


Human Motion Capture (Hasler etal. CVPR 2009)	Motion Deblurring (Cho etal. ICCV 2011)	Advanced Video Editing (Bhat etal. EGSR 2007)

Photometric Stereo (Joshi and Kriegman ICCV 2007)	Cloth Capture (Kavan etal. SIGGRAPH ASIA 2011)	Performance Capture for Editing (Jain etal. SIGGRAPH ASIA 2010)

Images courtesy of respective authors.

These are just a few examples, for a full list of topics see here.

Contact

In case you have questions about this seminar, please contact Levi Valgaerts – valgaerts (at) mpi-inf.mpg.de

Schedule

Date	What
April 17, 2012	First meeting of the seminar, presentation of topics. Compulsory for all Participants! (Slides)
May 10, 2012	Class room lecture: "How to give a good talk" (Slides) ATTENTION: TIME CHANGED! (Slides)
May 17, 2012	First seminar talk by participant. All other talks will follow in regular weekly slots.

Overview

The field of Computer Vision strives to develop algorithms for understanding, interpreting and reconstructing information of real world scenes from image and video data. The focus of Computer Graphics, on the other hand, is on image synthesis, i.e. on the development of algorithms to build and edit static and dynamic virtual worlds and to display them in photorealistic or stylised ways.

In recent years, it could be observed that these fields have more and more converged since both disciplines started to investigate research questions that are of relevance for either side. As an example, Computer Graphics researchers have started to investigate algorithms to reconstruct detailed models of static and dynamic scenes from image data, such that more believable virtual renderings of real world scenes could be achieved. In a similar manner, Computer Vision research has been able to benefit from the experience of Computer Graphics research to efficiently and effectively model image formation and light transport, which also simplifies scene analysis and reconstruction tasks. Recently, the ever increasing amount of image and video data has further strengthened the links between the two fields, as both communities start to investigate new ways to analyse, structure and immersively display these data.

This seminar covers advanced research topics that lie on the boundary of the fields of Computer Vision and Computer Graphics. These include classic and recent research results that were published at the top tier conferences and in top journals in either field. More specifically, the seminar will cover research papers from the following problem fields: Motion estimation and tracking, multi-view geometry and reconstruction, computational photography and videography, reconstruction of static and dynamic 3D scenes, advanced image and video processing

Target Audience

The target audience are graduate students in computer science or related fields. Basic knowledge in at least one of the fields of 3D geometry, computer vision, image processing, and computer graphics is required. Every participant will give a talk on a chosen scientific topic. Afterwards, the topic will be discussed within the seminar group. This is a great opportunity for students to improve their research and presentation skills and to learn about the latest developments in computer vision and computer graphics. The seminar language is English.

Structure

Part 1: Presentation

Each participant will perform a detailed study of one research topic by means of two representative scientific papers. He or she will present its main ideas in a presentation of approx. 45 min.

Work on the topic usually requires reading and understanding of the papers, as well as acquiring background information necessary to understand the topic. For extra background knowledge, some of the papers and books referenced in the papers may have to be consulted.

Your presentation should cover the main ideas of the chosen topic. Usually, this means that content has to be selected from the two representative papers, but it will often not be possible to discuss all results presented therein. Instead, try to focus on the topic and find a common thread linking the two papers on which you can build your presentation.

If you use formulas, make sure that all symbols are introduced properly. Similarly, make sure that figures are labelled correctly and that new terminology is introduced appropriately. You may assume that your audience is familiar with topics that have been presented in class earlier. It is very important for a good presentation to find a balance between overview and detail. Most importantly, the general theme of your topic should become clear from your presentation. Your goal is that the other participants will have gained a general understanding of your topic after your presentation. You can pick one or two things that you find interesting and present those in more detail. Finally, prepare your presentation on time and plan to spend some time on practising your talk. Our feedback will help you to improve it.

Important: We expect you to lead the discussion after your presentation. This requires that you prepare some questions, remarks or provocative statements for the audience. Try to challenge them. The discussion is also a good time to collect information and thoughts for the paper section where you present your own ideas. Keep this in mind when preparing your questions.

Part 2: Paper

In addition to the in-class presentation, we require a written report on the chosen topic. This report should summarise the main ideas of the accompanying papers and discuss limitations and drawbacks of the works. In addition, participants should develop and sketch an own idea of how to address one specific shortcoming. The report should consist of 6-8 pages covering the presented papers and about 2-3 pages covering the improvement proposed by the student.

The written report uses a template of a major computer vision conference, the IEEE Conference on Computer Vision and Pattern Recognition. Please use the review version of this template, as this makes it easier to refer to certain parts of your text. For your convenience we provide the template links here: .tar.gz (Linux) or .zip (Windows). The format of the written report has to be pdf. Please keep this in mind when using the Word template.

We will grade your report based on a number of criteria. As for the presentation, focus on getting the "big picture" and present your topic in context. Try to link the two articles in a coherent text, i.e. a structure like
Paper 1
Paper 2
is usually insufficient. Also, try to avoid copying the structure of the original papers, do not stick too closely to the original text and use your imagination to find a way of presenting the topic in your own words. Equations can of course be reproduced. The articles have been chosen such that they provide different views on a common theme. Try to exploit this.

Keep in mind to structure your paper well in a formal sense: An abstract provides a short overview. The introduction expands on that and gives an overview of your paper's structure. In a technical section you present your topic and the details that are of interest to you. Here you can also expand on the subject, i.e. add your own ideas. Finally, a conclusion summarises the paper. This is also the place where you can add your own opinions about the topic.

We encourage you to use additional literature in the development of your own ideas. We expect at least 3-4 additional references, if you want to include more, you are welcome to do so. Cite the references appropriately, e.g. if you use concepts from different publication or if you want to link to additional information. Your own ideas can be developed from scratch - if you have a great idea, try to sketch it and a possible solution. You can also discuss limitations or ideas that you feel are missing from the papers. A further possibility is to research recent developments following the papers of your topic. If you pursue this route, you could base your "own ideas" section on several follow-up publications, but make sure that you do not simply copy parts of other papers! A good resource for free links to research articles is Google Scholar. The "cited by" option often links to useful related work. Another source of freely available scientific articles is Citeseer. If you can only find a pay link, send us an email - we will try to get you a copy.

Your paper should be emailed to us by August 23, 2012.

Participation

Participants are expected to read every paper in preparation for the upcoming presentations. We expect students to actively engage in discussions for further understanding of the presented material. We aim at a creative atmosphere - ideas developed during the seminar work might lead to Master thesis projects. To encourage lively discussions, we decided that you can improve your grade by one step, e.g. from 2.0 to 1.7, if you participate actively in the discussion sessions after the talks.

Time and Place

Time:	Thursdays, 14:00-16:00 h
First Meeting:	Tuesday April 17, 2012, 17:00 h
Second Meeting:	Thursday May 10, 2012
Presentation Slots:	May 31, June 14, June 21, June 28, July 5, July 12, July 19, July 26
Room:	Building E1.4 (Max-Planck-Institut), Room 019

Organisation

Registration

Registration is open.

In case you want to participate in the seminar, send an email with your name, matriculation number, current semester and email address to valgaerts (at) mpi-inf.mpg.de. The details of the seminar and the available topics will be discussed in the first class on April 17. Attendance is required in all classes. If possible, take already a look at the list of topics and provide a ranked list of your three favorite topics: (1) best, (2) second best, etc. Based on your preference we plan to assign topics within one week. Please note that the number of available presentation slots is limited to ten. We will try to accommodate all requests for participation but in case we have more requests than available slots, we will assign slots on a first come first serve basis. All official announcements will be made through the mailing list. Please subscribe/unsubscribe to it.

Mailing List

itvc@lists.mpi-inf.mpg.de

Instructor Office Hours

Instructor	Time and Place	Contact
Ahmed	Monday, 11:00h - 12:00h	elhayek (at) mpi-inf.mpg.de
Miguel	Tuesday, 16:00h - 17:00h	granados (at) mpi-inf.mpg.de
Kiran	Thursday, 16:30h - 17:30h	varanasi (at) mpi-inf.mpg.de
Christian	TBA	theobalt (at) mpi-inf.mpg.de
Levi	Tuesday, 15:30h - 16:30h (only after appointment)	valgaerts (at) mpi-inf.mpg.de

List of Papers

Topic	Paper Title	Authors	Published at	Contact	Presenter	Date
3D Non-Rigid Motion Analysis	A continuous, editable representation for deforming mesh sequences with separate signals for time, pose and shape, (paper) Trajectory Space: A Dual Representation for Nonrigid Structure from Motion, (paper)	Thomas J. Cashman and Kai Hormann Ijaz Akhter, Yaser Sheikh, Sohaib Khan and Takeo Kanade	EUROGRAPHICS 2012 PAMI 2011	Kiran	Kiran	May 31	(slides)
Human Motion Capture	Fast Articulated Motion Tracking using a Sums of Gaussians Body Model, (paper) Markerless Motion Capture with Unsynchronized Moving Cameras, (paper)	Carsten Stoll, Nils Hasler, Juergen Gall, Hans-Peter Seidel and Christian Theobalt Nils Hasler, Bodo Rosenhahn, Thorsten Thormahlen, Michael Wand, Juergen Gall and Hans-Peter Seidel	ICCV 2011 CVPR 2009	Ahmed	Aurela Shehu	June 14	(Slides)
Video Stabilisation	Subspace Video Stabilization, (paper) Content-Preserving Warps for 3D Video Stabilization, (paper)	Feng Liu, Michael Gleicher, Jue Wang, Hailin Jin and Aseem Agarwala Feng Liu, Michael Gleicher, Hailin Jin and Aseem Agarwala	SIGGRAPH 2011 SIGGRAPH 2009	Miguel	Nadia Robertini	June 21	(Slides)
Image Matching	Data-driven Visual Similarity for Cross-domain Image Matching, (paper) Balanced Graph Matching, (paper)	Abhinav Shrivastava, Tomasz Malisiewicz, Abhinav Gupta and Alexei A. Efros Timothee Cour, Praveen Srinivasan and Jianbo Shi	SIGGRAPH Asia 2011 NIPS 2006	Kwang In	Kwang In	June 28	(Slides)
Scene Flow Estimation	Multi-View Scene Flow Estimation: A View Centered Variational Approach, (paper) Joint Estimation of Motion, Structure and Geometry from Stereo Sequences, (paper)	Tali Basha, Yael Moses, Nahum Kiryati Levi Valgaerts, Andres Bruhn, Henning Zimmer, Joachim Weickert, Carsten Stoll and Christian Theobalt	CVPR 2010 ECCV 2010	Levi	Elham AfsariYeganeh	July 5	(Slides)
~~Motion Capture with Depth Cameras~~	A Data-Driven Approach for Real-Time Full Body Pose Reconstruction from a Depth Camera, (paper) Real-Time Human Pose Recognition in Parts from Single Depth Images, (paper)	Andreas Baak, Meinard Müller, Gaurav Bharaj, Hans-Peter Seidel and Christian Theobalt Jamie Shotton, Andrew Fitzgibbon, Mat Cook, Toby Sharp, Mark Finocchio, Richard Moore, Alex Kipman and Andrew Blake	ICCV 2011 CVPR 2011	Christian	~~Maxim Reznitskii~~	~~July 12~~	–
Cloth Capture and Simulation of Advanced Scene Models	Physics-Inspired Upsampling for Cloth Simulation in Games, (paper) Video-based Reconstruction of Animatable Human Characters, (paper)	Ladislav Kavan, Dan Gerszewski, Adam W. Bargteil and Peter-Pike Sloan Carsten Stoll, Juergen Gall, Edilson de Aguiar, Sebastian Thrun and Christian Theobalt	SIGGRAPH Asia 2011 SIGGRAPH Asia 2010	Kiran	Artem Boldyrev	July 19	(Slides)
Passive Facial Performance Capture	High-Quality Passive Facial Performance Capture using Anchor Frames, (paper) High Resolution Passive Facial Performance Capture, (paper)	Thabo Beeler, Fabian Hahn, Derek Bradley, Bernd Bickel, Paul Beardsley, Craig Gotsman, Robert W. Sumner and Markus Gross Derek Bradley, Wolfgang Heidrich, Tiberiu Popa and Alla Sheffer	SIGGRAPH 2011 SIGGRAPH 2010	Levi	Levi	July 26	(Slides)