You are here: Home People David Hübner

David Hübner

 

huebner Brain State Decoding Lab
Albertstr. 23
D-79104 Freiburg im Breisgau 
Office: 00.009 (ground floor)


Phone:  +49 761 203 5330

 

Research Interests

My research has two main focus areas:

  1. Unsupervised machine learning for BCIs: The high subject-to-subject and session-to-session variability of brain signals requires a calibration session each time before using a BCI. Calibration is an arduous and time-consuming task. Even worse, it has been shown that the data distribution may change between calibration and online phase leading to suboptimal decoding performance.
    In my research, two unsupervised machine learning methods have been proposed to overcome these obstacles. First, we introduced Learning from Label Proportions (LLP) for BCIs. It utilizes the structure that is imposed by the paradigm to obtain a conceptually simple BCI system that does not require calibration and learns from unlabeled data and is guaranteed to obtain the optimal decoder. When combining it with an expectation-maximization (EM)-algorithm, we could show that unlabeled data is almost as useful as labeled data.
  2. Aphasia rehabilitation after stroke (see project CogReha): This interdisciplinary project has the goal to develop and evaluate a BCI-based training protocol for patients with aphasia (language deficits). A pilot study with 8 chronic patients already showed significant and strong improvements in various language tasks. We are currently preparing a larger control study.

Please find more information on my private homepage.

About me

  • Since Sep. 2015: PhD student in the BSDlab.
  • Sep. 2013 - Jul. 2015. Masters of Science in Computational Engineering and Applied Mathematics (Double degree program). TU Delft, Netherlands (1st year). KTH Stockholm, Sweden (2nd year). Degree Project in Computational Neuroscience.
  • Nov 2012 - Sep. 2013: Backpacking in Australia, New Zealand and various countries in South-East Asia.

  • Jul. 2012 - Nov. 2012: Study Abroad, The University of Western Australia, Perth, Australia. Various Subjects: Anthropology, Psychology, Finance.
  • Sep. 2009 - Jul. 2012: Bachelor of Science in Mathematics, University of Potsdam, Germany. Degree Project in Probability Theory about Markov Chains.

Publications

Journal Publications

Hübner D, Verhoeven T and Müller K-R, Kindermans P-J and Tangermann M (2018), "Unsupervised Learning for Brain-Computer Interfaces Based on Event-Related Potentials: Review and Online Comparison", IEEE Computational Intelligence Magazine. Vol. 13(2), pp. 66-77. IEEE. [ IEEE Xplore Homepage ] [pdf ]

Hübner D, Verhoeven T, Schmid K, Müller K-R, Tangermann M and Kindermans P-J (2017), "Learning from label proportions in brain-computer interfaces: Online unsupervised learning with guarantees", PLOS ONE. Vol. 12(4), pp. e0175856. Public Library of Science. [ pdf ]

Verhoeven T, Hübner D, Tangermann M, Müller K-R, Dambre J and Kindermans P-J (2017), "Improving zero-training brain-computer interfaces by mixing model estimators", Journal of Neural Engineering. Vol. 14(3), pp. 036021. IOP Publishing. [ JNE Homepage ]

Conference Publications

Hübner D, Verhoeven T, Kindermans P-J and Tangermann M (2017), "Mixing two unsupervised estimators for event-related potential decoding: An online evaluation", In Proceedings of the 7th International Brain-Computer Interface Meeting 2017: From Vision to Reality. pp. 198-203. Verlag der Technischen Universität Graz. Best Talk Award. [pdf]

Hübner D, Kindermans P-J, Verhoeven T and Tangermann M (2017), "Improving learning from label proportions by reducing the feature dimensionality", In Proceedings of the 7th International Brain-Computer Interface Meeting 2017: From Vision to Reality. pp. 186-191. Verlag der Technischen Universität Graz. [pdf]

Hübner D and Tangermann M (2017), "Challenging the assumption that auditory event-related potentials are independent and identically distributed", In Proceedings of the 7th International Brain-Computer Interface Meeting 2017: From Vision to Reality. pp. 192-197. Verlag der Technischen Universität Graz. [pdf]

Hübner D, Verhoeven T, Schmid K, Müller K-R, Tangermann M and Kindermans P-J (2017), "Learning from label proportions in BCI -- A symbiotic design for stimulus presentation and signal decoding", In The First Biannual Neuroadaptive Technology Conference. pp. 27-29. Best Talk Award. pdf ] [ slides ]

Kindermans P-J, Hübner D, Verhoeven T, Schmid K, Müller K-R and Tangermann M (2016), "Making Brain-Computer Interfaces robust, reliable and adaptive with Learning from Label Proportions", NIPS Workshop. [pdf]

Musso M, Bambadian A, Denzer S, Umarova R, Hübner D; and Tangermann M (2016), "A novel BCI based rehabilitation approach for aphasia rehabilitation", In Proceedings of the Sixth International Brain-Computer Interface Meeting: BCI Past, Present and Future. pp. 104.[ pdf ]