SlobodanIlic

Chair for Computer Aided Medical Procedures & Augmented Reality
Lehrstuhl für Informatikanwendungen in der Medizin & Augmented Reality

Slobodan Ilic

News

Me Dr. Slobodan Ilic
E-Mail:
Phone: +49 089-289 17082
Skype: Slobodan.Ilic.jf
Address: Technische Universität München
Fakultät für Informatik, I-16
Boltzmannstr. 3
85748 Garching b. München
Germany
Room: Room:03.13.040

Resume

Since February 2009 Slobodan Ilic is a leader of the Computer Vision Group of the CAMP Laboratorie at TUM, Germany. Form June 2006 he was a senior researcher at Deutsche Telekom Laboratories Deutsche Telekom Laboratories in Berlin. Before that he was a postdoctoral fellow for one year at Computer Vision Laboratory , EPFL, Switzerland, where he received his PhD in 2005. His research interests include mode-based Computer Vision, 3D modeling and reconstruction, and tracking of rigid and deformable objects.

Research Topics

I am currently involved in conducting research in the following research areas.

  • 3D deformable object modeling, shape recovery and tracking in monocular videos.
  • Tracking 3D surface deformations from multi-view videos (4D modeling).
  • Real-time 3D object detection and tracking.
  • Real-time motion-stereo and stereo fusion.
  • 3D object(rigid and deformable) detection in depth data.
  • Real-time template tracking.
  • Human-pose estimation.
  • Segmentation.

Computer Vision Group Team

Computer Vision Group consists of internal and external PhD students. External PhD student are financed by our industrial partners, and work on commonly defined PhD topics. Currently I work and advise/co-advise the following PhD students.

PhD students:

Master and bachelor students:

  • Simon Sklenak
  • Anke Schwarz
  • Jan Kremer
  • Maike Forberg

Alumni

Past PhD students

Past Master Students and Interns

  • David Tan
  • Vasilis Belagianis
  • Vladimir Haltakov
  • Hagen Kaiser -- SIEMENS and LMU
  • Benoit Diotte -- TUM

Teaching

I teach the following courses:

I would like to thank all authors whose slides I used for the lecture in tracking and detection. I cited most of them on the slides, but there are still some I didn't, so please find this statement as a replacement for citations on the slides.

I previously thought the course "Selected Topics in Computer Vision" together with Vincent Lepetit at the EPFL, which was about deformable object tracking.

Research projects

Deformable Object Tracking

Deformable Object Tracking

We address the problem of visual tracking of arbitrary objects that undergo signifi cant scale and appearance changes. The classical tracking methods rely on the bounding box surrounding the target object. Regardless of the tracking approach, the use of bounding box quite often introduces background information. This information propagates in time and its accumulation quite often results in drift and tracking failure. This is particularly the case with the particle filtering approach that is often used for visual tracking. However, it always uses a bounding box around the object to compute features of the particle samples. Since this causes the drift, we propose to use segmentation for sampling. Relying on segmentation and computing the colour and gradient orientation histograms from these segmented particle samples allows the tracker to easily adapt to the object's deformations, occlusions, orientation, scale and appearance changes. We propose two particle sampling strategies based on segmentation. In the rest, segmentation is done for every propagated particle sample, while in the second only the strongest particle sample is segmented. Depending on this decision there is obviously a trade-o between speed and performance. This project comes from a collaboration with EADS Innovations Works, Germany.
Linear Predictors

Linear Predictors

We address the problem of fast and robust tracking of free-form templates. For this we use Linear Predictors, introduced by Jurie & Dhome in 'Hyperplane Approximation for Template Matching', 2002. This approach allows to track templates at high frame-rates (>1000Hz) very robustly. However, the involved learning requires a significant amount of time and is therefore not suitable for many tasks where the environment is not known a-priori. Within this project we introduced several methods addressing this problem, starting with an adaptive approach that starts with learning a small template and then iteratively grows it over time. This approach also allows to adapt the size and shape of the tracked template during tracking, making it possible to handle occlusions. To speed-up the direct learning of large templates, we introduced two different approaches, a reformulation of the learning equations and a dimensionality reduction step. Both resulting in speed-ups up to two orders of magnitude.
Spatio Temporal Modeling of Dynamic Scenes

Spatio Temporal Modeling of Dynamic Scenes

A great part of Computer Vision research has been dedicated to shape recovery, tracking and detection of 3D objects in images and videos. While excellent results have been achieved in these areas, the majority of the methods still assume static scenes and rigid objects, and rarely explore temporal information. However, the world surrounding us is highly dynamic, and in many situations objects deform over time. This temporal information provides a richer and denser source of information and have not yet been extensively exploited. Our objective within this project is to explore spatio-temporal information in order to recover 3D shapes and the motion of the deformable objects. Therefore, we refer to this area as spatio-temporal or four dimensional modeling (4D modeling). With the increased popularity of 3D content in film industry, TV, Internet and games, tools and methods that exploit spatio-temporal information and allow fast and automated 3D content production are going to be indispensable.
Rigid 3D Object Detection

Rigid 3D Object Detection

Object detection and localization is a crucial step for inspection and manipulation tasks in robotic and industrial applications. We present an object detection and localization scheme for 3D objects that combines intensity and depth data. A novel multimodal, scale- and rotation-invariant feature is used to simultaneously describe the object's silhouette and surface appearance. The object's position is determined by matching scene and model features via a Hough-like local voting scheme. The proposed method is quantitatively and qualitatively evaluated on a large number of real sequences, proving that it is generic and highly robust to occlusions and clutter. Comparisons with state of the art methods demonstrate comparable results and higher robustness with respect to occlusions.
Stereo Vision

Stereo Vision

This project focuses on the development of novel methods for efficient or accurate binocular stereo vision.

Publications

Software error:

/srv/www/htdocs/publications//pub/: Input/output error at /srv/www/cgi-bin/pub/search.pl line 17.

For help, please send mail to the webmaster (/Main/MartinHorn), giving this error message and the time and date of the error.

For my previous publications please visit my old web page at Deutsche Telekom Laboratories or the publication page of CVLAB, EPFL.


UsersForm
Title: Dr.
Circumference of your head (in cm):  
Firstname: Slobodan
Middlename:  
Lastname: Ilic
Picture: photo.jpg
Birthday:  
Nationality: Serbia
Languages: English, German, French, Serbian
Groups: Computer Vision
Expertise: Computer Vision
Position: Scientific Staff
Status: Active
Emailbefore: slobodan.ilic
Emailafter: in.tum.de
Room: 03.13.041
Telephone: 089-289 17082
Alumniactivity:  
Defensedate:  
Thesistitle:  
Alumnihomepage:  
Personalvideo01:  
Personalvideotext01:  
Personalvideopreview01:  
Personalvideo02:  
Personalvideotext02:  
Personalvideopreview02:  


Edit | Attach | Refresh | Diffs | More | Revision r1.80 - 29 Apr 2013 - 09:58 - SlobodanIlic

Lehrstuhl für Computer Aided Medical Procedures & Augmented Reality    rss.gif