Quick Links

Contact

Dr. Stefan Harmeling

Address:	Spemannstr. 38 72076 Tübingen
Room number:	207
Phone:	+49 7071 601 1556
Fax:	+49 7071 601 552
E-Mail:	stefan.harmeling

Stefan Harmeling

Position: Group Leader Unit: Schölkopf

Overview
Publications

News:

My Google scholar page with citation counts
NIPS 2011 workshop on "Machine Learning meets Computational Photography (2)"
organized by Michael Hirsch, Stefan Harmeling, Rob Fergus (NYU), Peyman Milanfar (UCSC)
videos of all talks now on videolectures.net
NIPS 2011 workshop on "Cosmology meets Machine Learning"
organized by Sarah Bridle (UCL), Phil Marshall (UCL), Michael Hirsch, Stefan Harmeling, Mark Girolami (UCL), Bernhard Schülkopf
videos of all talks now on videolectures.net
video of invited talk on "Efficient space-variant blind deconvolution" at NIPS 2010 workshop on Numerical Mathematics Challenges in Machine Learning
NIPS 2010 workshop on "Machine Learning meets Computational Photography"
organized by Stefan Harmeling, Michael Hirsch, Peyman Milanfar (UCSC), Bill Freeman (MIT)
videos of all talks now on videolectures.net

Primary research interest:

Machine learning

Current research focus:

Machine learning applied to computational imaging

Besides recent publications on computational imaging, please find below selected publications on

supervised learning, unsupervised learning, and reinforcement learning.

Research on computational imaging

	*Hirsch, M., S. Harmeling, S. Sra and B. Schölkopf:* Online Multi-frame Blind Deconvolution with Super-resolution and Saturation Correction. Astronomy and Astrophysics xx, xx (accepted) (10 2010) Abstract
	*Harmeling, S., M. Hirsch and B. Schölkopf:* Space-Variant Single-Image Blind Deconvolution for Removing Camera Shake. Proceedings of the Twenty-Fourth Annual Conference on Neural Information Processing Systems (NIPS 2010) (accepted) (12 2010) Link
	*Harmeling, S., S. Sra, M. Hirsch and B. Schölkopf:* Multiframe Blind Deconvolution, Super-Resolution, and Saturation Correction via Incremental EM. Proceedings of the 17th International Conference on Image Processing (ICIP 2010) (accepted) (09 2010) Link
	*Hirsch, M., S. Sra, B. Schölkopf and S. Harmeling:* Efficient Filter Flow for Space-Variant Multiframe Blind Deconvolution. Proceedings of the Twenty-Third IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2010), 607-614 (06 2010) Abstract PDF Link DOI
	*Harmeling, S., M. Hirsch, S. Sra and B. Schölkopf:* Online blind deconvolution for Astronomy. Proceedings of the First IEEE International Conference Computational Photography (ICCP 2009), 1-7 (04 2009) Abstract PDF Link *code (mbd-0.0.tar)*

Research on supervised learning

	Bridle, S., S. T. Balan, M. Bethge, M. Gentile, S. Harmeling, C. Heymans, M. Hirsch, R. Hosseini, M. Jarvis, D. Kirk, T. Kitching, K. Kuijken, A. Lewis, S. Paulin-Henriksson, B. Schölkopf, M. Velander, L. Voigt, D. Witherick, A. Amara, G. Bernstein, F. Courbin, M. Gill and A. He: Results of the GREAT08 Challenge: An image analysis competition for cosmological lensing. Monthly Notices of the Royal Astronomical Society 405(3), 2044-2061 (07 2010) Abstract Link DOI
	*Baehrens, D., T. Schroeter, S. Harmeling, M. Kawanabe, K. Hansen and K.-R. Müller:* How to Explain Individual Classiﬁcation Decisions. Journal of Machine Learning Research 11, 1803-1831 (06 2010) Abstract PDF Link
	*Harmeling, S.:* Inferring textual entailment with a probabilistically sound calculus. Natural Language Engineering 15(4), 459-477 (10 2009) Abstract PDF Link DOI
	*Lampert, C. H., H. Nickisch and S. Harmeling:* Learning To Detect Unseen Object Classes by Between-Class Attribute Transfer. Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR 2009), 951-958, IEEE Service Center, Piscataway, NJ, USA (06 2009) Abstract PDF Link DOI

Research on unsupervised learning

	*Harmeling, S. and C. K. Williams:* Greedy Learning of Binary Latent Trees. IEEE Transactions on Pattern Analysis and Machine Intelligence Epub ahead (in press) (09 2010) Abstract Link DOI *code (ltt-1.4.tar)*
	*Harmeling, S.:* Exploring model selection techniques for nonlinear dimensionality reduction. Technical report No.(EDI-INF-RR-0960) (03 2007) Abstract PDF Link *code (nldim-0.1.tar)* (generates different types of swiss rolls, fishbowls, and much more)
	*Harmeling, S., G. Dornhege, D. Tax, F. Meinecke and K.-R. Müller:* From outliers to prototypes: Ordering data. Neurocomputing 69(13-15), 1608-1618 (08 2006) Abstract PDF Link DOI
	*Meinecke, F., S. Harmeling and K.-R. Müller:* Inlier-based ICA with an application to superimposed images. International Journal of Imaging Systems and Technology 15(1), 48-55 (07 2005) Abstract Link DOI *Meinecke, F., S. Harmeling and K.-R. Müller:* Robust ICA for Super-Gaussian Sources. Independent Component Analysis and Blind Signal Separation: Fifth International Conference (ICA 2004), 217-224. (Eds.) Puntonet, C. G., A. Prieto, Springer, Berlin, Germany (10 2004) Abstract DOI *code (ibica-0.1.tar)*
	*Harmeling, S., F. Meinecke and K.-R. Müller:* Injecting noise for analysing the stability of ICA components. Signal Processing 84(2), 255-266 (02 2004) Abstract Link DOI *Harmeling, S., F. Meinecke and K.-R. Müller:* Analysing ICA component by injection noise. Proceedings of the 4th International Symposium on Independent Component Analysis and Blind Signal Separation (ICA 2003), 149-154. (Eds.) Amari, S.-I., A. Cichocki, S. Makino, N. Murata (04 2003) Abstract PDF Link *code (relica-0.1.tar)*
	*Harmeling, S., A. Ziehe, M. Kawanabe and K.-R. Müller:* Kernel-based nonlinear blind source separation. Neural Computation 15(5), 1089-1124 (05 2003) Abstract PDF Link DOI *Harmeling, S., A. Ziehe, M. Kawanabe and K.-R. Müller:* Kernel feature spaces and nonlinear blind source separation. Advances in Neural Information Processing Systems 14: Proceedings of the NIPS 2001 Conference, 761-768. (Eds.) Dietterich, T. G., S. Becker, Z. Ghahramani, MIT Press, Cambridge, MA, USA (09 2002) Abstract PDF Link *Harmeling, S., A. Ziehe, M. Kawanabe, B. Blankertz and K.-R. Müller:* Nonlinear blind source separation using kernel feature spaces. Proceedings of the Third International Workshop on Independent Component Analysis and Blind Signal Separation (ICA 2001), 102-107. (Eds.) Lee, T.-W., T.P. Jung, S. Makeig, T. J. Sejnowski (12 2001) Abstract PDF *code (ktdsep-0.2.tar)*

Research on reinforcement learning

Toussaint, M., A. Storkey and S. Harmeling: Expectation-Maximization methods for solving (PO)MDPs and optimal control problems. Inference and Learning in Dynamic Models, xx, Cambridge University Press, UK (in press) (2010)
PDF

Toussaint, M., S. Harmeling and A. Storkey: Probabilistic inference for solving (PO)MDPs. Technical report No.(934) (12 2006)
PDF

Miscellaneous code

Schölkopf/Smola/Müller's kernel principal component analysis (KPCA)
Matlab-Implemenation, for more information see Bernhard Schölkopf, Alexander Smola, Klaus-Robert Müller, Nonlinear component analysis as a kernel eigenvalue problem, Neural Computation 10, 5, 1299-1319, 1996.
code (kpca-0.1.tar)

Parra/Spence's blind source separation algorithm
Matlab-Implemenation
code (convbss-0.1.tar)

Genetic Algorithms applied to 3-SAT problems
C Programming Language
code (gasat-1.1.tar)

Articles (15):
•	Zscheischler J , Mahecha MD , Harmeling S and Reichstein M (2013) Detection and attribution of large spatiotemporal extreme events in Earth observation data Ecological Informatics 15 66-73.
•	Burger HC , Schuler CJ and Harmeling S (2012) Image denoising with multi-layer perceptrons, part 1: comparison with existing algorithms and with bounds . submitted
•	Burger HC , Schuler CJ and Harmeling S (2012) Image denoising with multi-layer perceptrons, part 2: training trade-offs and analysis of their mechanisms . submitted
•	Kitching T , Amara A , Gill M , Harmeling S , Heymans C , Massey R , Rowe B , Schrabback T , Voigt L , Balan S , Bernstein G , Bethge M , Bridle S , Courbin F , Gentile M , Heavens A , Hirsch M , Hosseini R , Kiessling A , Kirk D , Kuijken K , Mandelbaum R , Moghaddam B , Nurbaeva G , Paulin-Henriksson S , Rassat A , Rhodes J , Schölkopf B , Shawe-Taylor J , Shmakova M , Taylor A , Velander M , van Waerbeke L , Witherick D and Wittman D (2011) Gravitational Lensing Accuracy Testing 2010 (GREAT10) Challenge Handbook Annals of Applied Statistics 5(3) 2231-2263.
•	Hirsch M , Harmeling S , Sra S and Schölkopf B (2011) Online Multi-frame Blind Deconvolution with Super-resolution and Saturation Correction Astronomy & Astrophysics 531(A9) 11 pages.
•	Harmeling S and Williams CK (2011) Greedy Learning of Binary Latent Trees IEEE Transactions on Pattern Analysis and Machine Intelligence 33(6) 1087-1097.
•	Bridle S , Balan ST , Bethge M , Gentile M , Harmeling S , Heymans C , Hirsch M , Hosseini R , Jarvis M , Kirk D , Kitching T , Kuijken K , Lewis A , Paulin-Henriksson S , Schölkopf B , Velander M , Voigt L , Witherick D , Amara A , Bernstein G , Courbin F , Gill M , Heavens A , Mandelbaum R , Massey R , Moghaddam B , Rassat A , Refregier A , Rhodes J , Schrabback T , Shawe-Taylor J , Shmakova M , van Waerbeke L and Wittman D (2010) Results of the GREAT08 Challenge: An image analysis competition for cosmological lensing Monthly Notices of the Royal Astronomical Society 405(3) 2044-2061.
•	Baehrens D , Schroeter T , Harmeling S , Kawanabe M , Hansen K and Müller K-R (2010) How to Explain Individual Classification Decisions Journal of Machine Learning Research 11 1803-1831.
•	Harmeling S (2009) Inferring textual entailment with a probabilistically sound calculus Natural Language Engineering 15(4) 459-477.
•	Harmeling S , Dornhege G , Tax D , Meinecke F and Müller K-R (2006) From outliers to prototypes: Ordering data Neurocomputing 69(13-15) 1608-1618.
•	Meinecke F , Harmeling S and Müller K-R (2005) Inlier-based ICA with an application to superimposed images International Journal of Imaging Systems and Technology 15(1) 48-55.
•	Oja E , Harmeling S and Almeida L (2004) Independent component analysis and beyond Signal Processing 84(2) 215-216.
•	Harmeling S , Meinecke F and Müller K-R (2004) Injecting noise for analysing the stability of ICA components Signal Processing 84(2) 255-266.
•	Ziehe A , Kawanabe M , Harmeling S and Müller K-R (2003) Blind separation of post-nonlinear mixtures using linearizing transformations and temporal decorrelation Journal of Machine Learning Research 4(7-8) 1319-1338.
•	Harmeling S , Ziehe A , Kawanabe M and Müller K-R (2003) Kernel-based nonlinear blind source separation Neural Computation 15(5) 1089-1124.
Conference papers (26):
•	Köhler R , Hirsch M , Schölkopf B and Harmeling S (2013) Improving alpha matting and motion blurred foreground estimation IEEE Conference on Image Processing (ICIP 2013).
•	Schuler C , Burger HC , Harmeling S and Schölkopf B (2013) A machine learning approach for image deconvolution IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2013). accepted
•	Grosse-Wentrup M , Harmeling S , Zander T , Hill J and Schölkopf B (2013) How to Test the Quality of Reconstructed Sources in Independent Component Analysis (ICA) of EEG/MEG Data In: Proceedings of the 3rd International Workshop on Pattern Recognition in NeuroImaging (PRNI 2013), 3rd International Workshop on Pattern Recognition in NeuroImaging (PRNI 2013). accepted
•	Harmeling S , Hirsch M and Schölkopf B (2013) On a link between kernel mean maps and Fraunhofer diffraction, with an application to super-resolution beyond the diffraction limit IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2013). accepted
•	Zscheischler J , Mahecha MD and Harmeling S (2012) Climate classifications: the value of unsupervised clustering In: Procedia Computer Science Volume 9, (Ed) H. Ali, Y. Shi, D. Khazanchi, M. Lees, G.D. van Albada, J. Dongarra, P.M.A. Sloot, J. Dongarra, International Conference on Computational Science (ICCS 2012), Elsevier, Amsterdam, Netherlands, 897-906.
•	Burger HC , Schuler CJ and Harmeling S (2012) Image denoising: Can plain Neural Networks compete with BM3D? 25th IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2012), 2392 - 2399.
•	Schuler CJ , Hirsch M , Harmeling S and Schölkopf B (2012) Blind Correction of Optical Aberrations In: Computer Vision - ECCV 2012, LNCS Vol. 7574, (Ed) S Lazebnik A Fitzgibbon and C Schmid, 12th IEEE European Conference on Computer Vision (ECCV 2012), Springer, Berlin, Germany, 187-200.
•	Köhler R , Hirsch M , Mohler B , Schölkopf B and Harmeling S (2012) Recording and Playback of Camera Shake: Benchmarking Blind Deconvolution with a Real-World Database In: Computer Vision - ECCV 2012, LNCS Vol. 7578, (Ed) A Fitzgibbon, S Lazebnik, P Perona, Y Sato and C Schmid, 12th European Conference on Computer Vision (ECCV 2012), Springer, Berlin, Germany, 27-40.
•	Hirsch M , Schuler CJ , Harmeling S and Schölkopf B (2011) Fast removal of non-uniform camera shake (Ed) DN Metaxas, L Quan, A Sanfeliu and LJ Van Gool, 13th IEEE International Conference on Computer Vision (ICCV 2011), IEEE, Piscataway, NJ, USA, 463-470.
•	Schuler CJ , Hirsch M , Harmeling S and Schölkopf B (2011) Non-stationary correction of optical aberrations (Ed) DN Metaxas, L Quan, A Sanfeliu and LJ Van Gool, 13th IEEE International Conference on Computer Vision (ICCV 2011), IEEE, Piscataway, NJ, USA, 659-666.
•	Loktyushin A and Harmeling S (2011) Automatic foreground-background refocusing (Ed) Macq, B. , P. Schelkens, 18th IEEE International Conference on Image Processing (ICIP 2011), IEEE, Piscataway, NJ, USA, 3445-3448.
•	Burger HC and Harmeling S (2011) Improving Denoising Algorithms via a Multi-scale Meta-procedure In: Pattern Recognition, (Ed) Mester, R. , M. Felsberg, 33rd DAGM Symposium, Springer, Berlin, Germany, 206-215.
•	Burger HC , Schölkopf B and Harmeling S (2011) Removing noise from astronomical images using a pixel-specific noise model (Ed) H Lensch, SL Narasimhan and ME Testorf, IEEE International Conference on Computational Photography (ICCP 2011), IEEE, Piscataway, NJ, USA, 8 pages.
•	Harmeling S , Sra S , Hirsch M and Schölkopf B (2010) Multiframe Blind Deconvolution, Super-Resolution, and Saturation Correction via Incremental EM 17th International Conference on Image Processing (ICIP 2010), IEEE, Piscataway, NJ, USA, 3313-3316.
•	Hirsch M , Sra S , Schölkopf B and Harmeling S (2010) Efficient Filter Flow for Space-Variant Multiframe Blind Deconvolution In: Proceedings of the 23rd IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2010, IEEE, Piscataway, NJ, USA, 607-614.
•	Harmeling S , Hirsch M and Schölkopf B (2010) Space-Variant Single-Image Blind Deconvolution for Removing Camera Shake In: Advances in Neural Information Processing Systems 23, (Ed) J Lafferty, CKI Williams, J Shawe-Taylor, RS Zemel and A Culotta, 24th Annual Conference on Neural Information Processing Systems (NIPS 2010), Curran, Red Hook, NY, USA, 829-837.
•	Lampert CH , Nickisch H and Harmeling S (2009) Learning To Detect Unseen Object Classes by Between-Class Attribute Transfer In: CVPR 2009, IEEE Computer Society Conference on Computer Vision and Pattern Recognition, IEEE Service Center, Piscataway, NJ, USA, 951-958.
•	Harmeling S , Hirsch M , Sra S and Schölkopf B (2009) Online blind deconvolution for astronomical imaging First IEEE International Conference on Computational Photography (ICCP 2009), IEEE, Piscataway, NJ, USA, 1-7.
•	Harmeling S (2007) An Extensible Probabilistic Transformation-based Approach to the Third Recognizing Textual Entailment Challenge In: TextEntail 2007, ACL-PASCAL Workshop on Textual Entailment and Paraphrasing, 137-142.
•	Meinecke F , Harmeling S and Müller K-R (2004) Robust ICA for Super-Gaussian Sources In: ICA 2004, (Ed) Puntonet, C. G., A. Prieto, Fifth International Conference on Independent Component Analysis and Blind Signal Separation, Springer, Berlin, Germany, 217-224.
•	Honkela A , Harmeling S , Lundqvist L and Valpola H (2004) Using kernel PCA for Initialisation of Variational Bayesian Nonlinear Blind Source Separation Method In: ICA 2004, (Ed) Puntonet, C. G., A. Prieto, Fifth International Conference on Independent Component Analysis and Blind Signal Separation, Springer, Berlin, Germany, 790-797.
•	Harmeling S , Meinecke F and Müller K-R (2003) Analysing ICA component by injection noise In: ICA 2003, (Ed) Amari, S.-I. , A. Cichocki, S. Makino, N. Murata, 4th International Symposium on Independent Component Analysis and Blind Signal Separation, 149-154.
•	Ziehe A , Kawanabe M , Harmeling S and Müller K-R (2003) Blind separation of post-nonlinear mixtures using gaussianizing transformations and temporal decorrelation In: ICA 2003, (Ed) Amari, S.-I. , A. Cichocki, S. Makino, N. Murata, 4th International Symposium on Independent Component Analysis and Blind Signal Separation, 269-274.
•	Harmeling S , Ziehe A , Kawanabe M and Müller K-R (2002) Kernel feature spaces and nonlinear blind source separation In: Advances in Neural Information Processing Systems 14, (Ed) Dietterich, T. G., S. Becker, Z. Ghahramani, Fifteenth Annual Neural Information Processing Systems Conference (NIPS 2001), MIT Press, Cambridge, MA, USA, 761-768.
•	Harmeling S , Ziehe A , Kawanabe M , Blankertz B and Müller K-R (2001) Nonlinear blind source separation using kernel feature spaces In: ICA 2001, (Ed) Lee, T.-W. , T.P. Jung, S. Makeig, T. J. Sejnowski, Third International Workshop on Independent Component Analysis and Blind Signal Separation, 102-107.
•	Ziehe A , Kawanabe M , Harmeling S and Müller K-R (2001) Separation of post-nonlinear mixtures using ACE and temporal decorrelation In: ICA 2001, (Ed) Lee, T.-W. , T.P. Jung, S. Makeig, T. J. Sejnowski, Third International Workshop on Independent Component Analysis and Blind Signal Separation, 433-438.
Contributions to books (2):
•	Toussaint M , Storkey A and Harmeling S : Expectation-Maximization methods for solving (PO)MDPs and optimal control problems, -. In: Inference and Learning in Dynamic Models, (Ed) Barber, D. , A.T. Cemgil, S. Chiappa, Cambridge University Press, Cambridge, UK, (2012). in press
•	Harmeling S : Solving Satisfiability Problems with Genetic Algorithms, 206-213. In: Genetic Algorithms and Genetic Programming at Stanford 2000, (Ed) Koza, J. R., Stanford Bookstore, Stanford, CA, USA, (2000).
Technical reports (9):
•	Schuler CJ , Hirsch M , Harmeling S and Schölkopf B : Non-stationary Correction of Optical Aberrations, 1, Max Planck Institute for Intelligent Systems, Tübingen, Germany, (2011).
•	Harmeling S , Sra S , Hirsch M and Schölkopf B : An Incremental GEM Framework for Multiframe Blind Deconvolution, Super-Resolution, and Saturation Correction, 187, Max Planck Institute for Biological Cybernetics, Tübingen, Germany, (2009).
•	Hirsch M , Sra S , Schölkopf B and Harmeling S : Efficient Filter Flow for Space-Variant Multiframe Blind Deconvolution, 188, Max Planck Institute for Biological Cybernetics, Tübingen, Germany, (2009).
•	Harmeling S and Toussaint M : Bayesian Estimators for Robins-Ritovs Problem, EDI-INF-RR-1189, School of Informatics, University of Edinburgh, (2007).
•	Harmeling S : Exploring model selection techniques for nonlinear dimensionality reduction, EDI-INF-RR-0960, School of Informatics, University of Edinburgh, (2007).
•	Toussaint M , Harmeling S and Storkey A : Probabilistic inference for solving (PO)MDPs, 934, School of Informatics, University of Edinburgh, (2006).
•	Jutten C , Karhunen J , Almeida L and Harmeling S : Technical report on Separation methods for nonlinear mixtures, D29, EU-Project BLISS, (2003).