Navigation

You are here: Home / Members / Michael Donoser

Homepage of Donoser Michael [Last Update: August 2014]

Michael Donoser was a senior researcher with tenure track at the Institute for Computer Graphics and Vision at Graz University of Technology. He received a Doctor of Philosophy degree (PhD) in computer science in 2007, after finishing his Master of Science (MSc) in 2003 all at Graz University of Technology, Austria. From 2007 until end of 2012 he was leading a group of young, motivated PhDs, the Virtual Habitat Group (see link below), initiating and supervising the fruitful ICG-Reading Group (see link below). From 2012 until July 2014 he worked as a senior researcher at ICG. In July 2014 he resigned from all his duties, becoming the manager of the computer vision team at Amazon, Berlin. If you are interested in publications, scientific topics and code - take a look at the Virtual Habitat page (see link)

 

QUICK LINKS

Publications (PDFs)

Research Overview

Curriculum Vitae

Virtual Habitat Group

Reading Group

News

All current publications as PDFs can be found at my publication page.

  • August, 15th 2013: First day of work at Amazon, Berlin as senior computer vision scientist

    After 12 years at ICG, TU Graz I moved to Berlin to work as the manager of the computer vision team for the Amazon Development Center. We are currently hiring computer vision people at various levels of expertise, so if you are interested, please contact me at michael.donoser(at)gmail.com

  • July, 31th 2013: Last day of work after 12 years being part of ICG

    After 12 years at ICG, TU Graz I resigned my tenure track contract to become part of a big company. Thanks to all who made working at ICG so interesting, entertaining and fruitful - especially I thank all the PhD students that I met during my time at ICG.

  • March, 1th 2014: Two CVPR 2014 papers accepted

    The papers "Discriminative Feature to-Point Matching for Image-Based Localization" by Michael Donoser and Dieter Schmalstieg and "Discrete-Continuous Gradient Orientation Estimation for Faster Image Segmentation" by Michael Donoser and Dieter Schmalstieg have been accepted for publication at the Conference on Computer Vision and Pattern Recognition (CVPR) 2014.

  • March, 1th 2014: Two BMVC 2013 papers accepted (1 oral and 1 poster)

    The papers "Replicator Graph Clustering" by Michael Donoser and "Incremental Surface Extraction from Sparse Structure-from-Motion Point Clouds" by Christof Hoppe, Manfred Klopschitz, Michael Donoser and Horst Bischof have been accepted for publication at the British Machine Vision Conference (BMVC) 2013.

  • March, 4th 2013: Two CVPR 2013 papers accepted (1 Oral + 1 Poster)

    The papers "Diffusion Processes for Retrieval Revisited" (ORAL) by Michael Donoser and Horst Bischof and "Optimizing 1-Nearest Prototype Classifiers" (POSTER) by Paul Wohlhart, Martin Köstinger, Michael Donoser, Peter Roth and Horst Bischof were accepted to CVPR 2013.

  • November, 15th 2012: Best Paper Award at ACCV 2012

    The paper "Detecting Partially Occluded Objects with an Implicit Shape Model Random Field" by Paul Wohlhart, Michael Donoser, Peter M. Roth and Horst Bischof we awarded the Best Scientific Paper Award at the Asian Conference on Computer Vision, 2012 in Daejeon, Korea.

  • June, 20th 2012: ECCV 2012 paper accepted

    The paper "Hough Regions for Joining Instance Localization and Segmentation" by Riemenschneider, Sternig, Donoser, Roth and Bischof was accepted to the European Conference on Computer Vision (ECCV), 2012.

  • June, 15th 2012: BMVC 2012 paper accepted

    The paper "Learning Edge-Specific Kernel Functions For Pairwise Graph Matching" by Donoser, Urschler and Bischof was accepted to the British Machine Vision Conference (BMVC), 2012.

  • March, 15th 2012: CVPR 2012 paper accepted

    The paper "Irregular lattices for complex shape grammar facade parsing" by Riemenschneider, Krispel, Thaller, Donoser, Havemann, Fellner and Bischof was accepted to the Conference on Computer Vision and Pattern Recognition (CVPR) 2012.

  • September, 10th 2011: ISMAR 2011 paper accepted

    The paper "Robust Planar Target Tracking and Pose Estimation from a Single Concavity" from Michael Donoser, Peter Kontschieder and Horst Bischof, was accepted as a full paper at ISMAR 2011.

  • June, 28th 2011: Two BMVC 2011 papers accepted

    The papers "Discriminative Learning of Contour Fragments for Object Detection" (ORAL) by P. Kontschieder, H. Riemenschneider, M. Donoser and H. Bischof and "Image Labelling as a Label Puzzle Game" (POSTER) by P. Kontschieder, S. Rota Bulò, M. Donoser, M. Pelillo and Bischof, H. were accepted to the British Machine Vision Conference (BMVC), 2011.

  • June, 10th 2010: ECCV 2010 paper accepted

    The paper "Using Partial Edge Contour Matches for Efficient Object Category Localization" by Hayko Riemenschneider, Michael Donoser and Horst Bischof was accepted for presentation at the European Conference on Computer Vision (ECCV) in Crete, Greece.

  • May, 31th 2010: Best Scientific Paper Award at OAGM 2010

    The paper "MSER Templates for 3D Pose Tracking" by Peter Kontschieder, Michael Donoser and Horst Bischof was awarded with the Best Scientific Paper award at the Workshop of the Austrian Association for Pattern Recognition

  • February, 3th 2010: CVPR 2010 paper accepted

    The paper "Linked Edges as Stable Region Boundaries" from Donoser, M., Riemenschneider, H. and Bischof, H. was accepted to CVPR 2010.

  • July, 15th 2009: Two ACCV 2009 papers and one BMVC 2009 paper accepted

    The papers "Efficient Partial Shape Matching of Outer Contours" from Donoser, M., Riemenschneider, H. and Bischof, H. ACCV 2009 and "Beyond Pairwise Shape Similarity Analysis" from Kontschieder, P., Donoser, M. and Bischof, H. (Poster) ACCV 2009 and "Bag of Optical Flow Volumes for Image Sequence Recognition" from Riemenschneider, H., Donoser, M. and Bischof, H., (Poster) BMVC 2009 were accepted.

  • June, 10th 2009: ICCV 2009 paper accepted

    The paper "Saliency Driven Total Variational Segmentation by Michael Donoser, Martin Urschler, Martin Hirzer and Horst Bischof was accepted to the International Conference on Computer Vision (ICCV) 2009 (PDF)

  • May, 14th 2009: Best Scientific Paper Award at OAGM 2009

    The paper "Finding Stable Extremal Region Boundaries" from Hayko Riemenschneider, Michael Donoser, and Horst Bischof was awarded with the Best Scientific Paper award at the Workshop of the Austrian Association for Pattern Recognition

  • December, 10th 2008: Best Scientific Paper Award at ICPR 2008

    Our paper "Using Web Search Engines to Improve Text Recognition" from Donoser M., Wagner, S. and Bischof, H. was awarded with the Best Scientific Paper awarde at the International Conference on Pattern Recognition (ICPR 2008) in Tampa Bay, Florida. A picture can be found here.

  • September, 3rd 2008: Oral Presentation at BMVC 2008

    Oral presentation of the paper "Fast Non-Rigid Object Boundary Tracking", M. Donoser and H. Bischof at the British Machine Vision Conference (BMVC 2008) in Leeds, UK.

  • May, 12th 2008: Award of State Styria

    I got the Josef-Krainer 2008-Award of the state styria for outstanding scientific work in the field of computer vision. A picture can be found here.

Curriculum Vitae

You can find my CV including a publication list here [Last Update: March, 2013].

PhD

My Phd Thesis on 3D Paper Structure Analysis using Computer Vision was finished in April 2007.

Lectures

I was lecturer of several fundamental and computer vision-specific courses like Introduction to Scientific Working, Pattern Recognition Seminar, Selected Topics of Computer Vision and Introduction to Programming until January, 2013. Resources concerning these lectures are made available on demand. For recent information to these lectures, please check the TUG-Online.

Research

You can find a detailed summary of my main research topics (including CODE and EXAMPLES) at the VH research page. But in general, my main computer vision interest lies on complete scene understanding, i.e. given a single image, we want to explain the whole image, by simultaneously recognizing and localizing all category instances in an image and by identifying their accurate spatial extent, i.e. providing a pixel-level segmentation per recognized category instance. Obviously, such a goal involves many important computer vision steps like feature extraction and description, object category recognition and unsupervised segmentation and I aimed at contributing to all these fields in the last years.

Publications

A list of my publications can be found at my publication page. A comprehensive list of all publications (including all PDFs) from my Virtual Habitat group can be found at the Virtual Habitat page.

Document Actions

[Powered by Plone]