Dr.-Ing. Fabian Timm University of Lübeck Institute for Neuro- and Bioinformatics Ratzeburger Allee 160 (Geb. 64) D-23562 Lübeck Germany Tel:     +49 451 500 5513 Fax:    +49 451 500 5502 Email: timm (at) inb (dot) uni-luebeck (dot) de

Research Interests

Demos and Experiments

• Video on accurate eye centre localisation, find paper presented at VISAPP 2011 below
• Analysis of human eye movements for the inspection of "Waldo" images: preliminary results
  

Bachelor's / Master's / Diploma-Thesis 

Source Code and Datasets

• OneClassMaxMinOver (OMMO) for training a one-class support vector machine [matlab code]
• Artificial images for the detection of round objects
  

Publications

1. Fabian Timm and Erhardt Barth. Novelty Detection for the Inspection of Light-emitting Diodes. Expert Systems with Applications. 2011. (accepted)
   
2. Fabian Timm and Erhardt Barth. Accurate eye centre localisation by means of gradients. In Proceedings of the Int. Conference on Computer Vision Theory and Applications (VISAPP), volume 1, pages 125-130, Algarve, Portugal, 2011. INSTICC. [demo video] [pdf]
   
3. Sascha Klement, Fabian Timm and Erhardt Barth. Illumination Correction for Image Stitching. In Proceedings of the Int. Conference on Imaging Theory and Applications (IMAGAPP), volume 1, pages 81-86, Algarve, Portugal, 2011. INSTICC. [pdf]
4. Fabian Timm and Erhardt Barth. Non-parametric texture defect detection using Weibull features. In Image Processing: Machine Vision Applications IV, San Francisco, USA, 2011. IS&T/SPIE Electronic Imaging. [pdf]
   
5. Fabian Timm and Erhardt Barth. Accurate, fast, and robust centre localisation for images of semiconductor components. In Image Processing: Machine Vision Applications IV, San Francisco, USA, 2011. IS&T/SPIE Electronic Imaging. [pdf]
   
6. Fabian Timm and Thomas Martinetz. Statistical Fourier Descriptors for Defect Image Classification. In Proceedings of the 20th Int. Conference on Pattern Recognition (ICPR), pages 4190-4193, Istanbul, Turkey, 2010. IEEE Computer Society Press. [pdf] 
7. Fabian Timm, Thomas Martinetz, and Erhardt Barth. Optical Inspection of Welding Seams. In Communications in Computer Science and Information Science, VISIGRAPP 2009 Revised Selected Papers, pages 269-282, Springer, 2010. [pdf]
8. Fabian Timm, Sascha Klement, Thomas Martinetz, and Erhardt Barth. Welding Inspection Using Novel Specularity Features and a One-class SVM. In Proceedings of the Int. Conference on Computer Vision Theory and Applications (VISAPP), volume 1, pages 146-153, Lisboa, Portugal, 2009. INSTICC. [pdf]
9. Kai Labusch, Fabian Timm, and Thomas Martinetz. Simple Incremental One-class Support Vector Classification. In Gerhard Rigoll, editor, Pattern Recognition - Proceedings of the DAGM, Lecture Notes in Computer Science, pages 21-30, 2008. [pdf] [matlab code]
10. Fabian Timm, Sascha Klement, and Thomas Martinetz. Fast Model Selection for MaxMinOver-based Training of Support Vector Machines. In Proceedings of the 19th Int. Conference on Pattern Recognition (ICPR), Tampa, Florida, USA, 2008. IEEE Computer Society Press. [pdf]
11. Fabian Timm. Methoden zur optischen Qualitätskontrolle. Anwendung in der Produktion von Xenon-Lampen. Diploma thesis, Universität zu Lübeck, 2006. Full version: Please contact the author via email. [pdf]

Curriculum vitae

• January 2012 - 
  Research & Development Engineer at Robert Bosch GmbH, Stuttgart, Germany.
• November 2011 - December 2011
  Post-Doc at the Institute for Neuro- and Bioinformatics, University of Lübeck, Germany.
• December 2006 - November 2011
  PhD student at the Institute for Neuro- and Bioinformatics, University of Lübeck, Germany.
  PhD thesis "Machine Vision for Inspection and Novelty Detection", submitted Jun 2011.
  
• December 2005 - October 2006:
  Diploma thesis "Methods for the Optical Inspection of Xenon-Lamps"
  At Philips Global Technology Development Aachen (Lighting)
• October 2001 - November 2006:
  Student of Computer Science, Bioinformatics, and Biomathematics at the University of Lübeck. Graduated with the degree of Dipl.-Inf. (What is a Dipl.-Inf.?)