Julien Vitay

Lecturer at the Chemnitz University of Technology, in the lab of Artificial Intelligence of the Faculty of Computer Science.

Previously postdoc in the Psychology Department of the University of Münster (Germany), under the supervision of Prof. Dr. Fred Hamker and PhD student at the Inria Nancy (Lorraine, France), in the CORTEX lab headed by Dr. Frédéric Alexandre.

Research themes

Artificial Intelligence

  • Decision-making
  • Goal-directed learning
  • Emotions and Motivation
  • Working memory
  • Reinforcement learning
  • Interval timing
  • Autonomous robotics

Computational Neuroscience

  • Basal Ganglia
  • Prefrontal cortex
  • Amygdala
  • Hippocampus
  • Afferents to the dopaminergic system (lateral habenula, pedunculopontine nucleus...)

Machine Learning

  • Deep neural networks
  • Reinforcement learning
  • Model compression
  • Support-vector machines

Parallel computing

  • Neural simulators
  • Code generation
  • OpenMP, MPI, CUDA
  • Auto-tuning

Curriculum Vitae

Born on December 11th, 1979 in Saint-Nazaire (Loire-Atlantique, France).


  • 2017 - Habilitation in Computer Science from the Chemnitz University of Technology (Saxony, Germany).
  • 2006 - PhD in Computer Science from the University Henri Poincaré Nancy-I (Lorraine, France).
  • 2002 - MSc in Microelectronics from the University of Rennes-I (Bretagne, France).
  • 2002 - Engineer grade from the École Supérieure d'Electricité (Supélec) in Rennes (Bretagne, France).
  • 1997 - Baccalauréat Scientifique, Lycée Galilée, Guérande (Loire-Atlantique, France).

Professional experience

  • 2011 - now : Lecturer at the Department of Computer Science, Chemnitz University of Technology (Saxony, Germany).
  • 2006 - 2011 : Postdoc at the Institute of Psychology, University of Münster (North Rhine-Westphalia, Germany).
  • 2005 - 2006 : Teaching assistant at the ESIAL engineering school of the University Henri Poincaré Nancy-I (Lorraine, France).
  • 2002 - 2005 : Teaching assistant at the Department of Computer Science, University Henri Poincaré Nancy-I (Lorraine, France).


  • Natural languages: French, English, German
  • Programming languages: Python, C++, C, Matlab, Java, Julia, VHDL
  • Parallel frameworks: OpenMP, MPI, CUDA
  • OS: GNU/Linux, MacOSX



  • Vitay, J. (2017). "On the role of dopamine in motivated behavior: a neuro-computational approach", Habilitation TU Chemnitz, Januar 2017.
  • Vitay, J. (2006). "Emergence de fonctions sensorimotrices sur un substrat neuronal numérique distribué", Doctorat de l'Université Henri-Poincaré Nancy-I, June 23rd 2006.

Peer-reviewed journals

Book chapters

  • Vitay, J. and Hamker, FH. (2012), "Basal Ganglia learning", in Encyclopedia of the Sciences of Learning, N.M. Seel (Ed.), Springer US.
  • Vitay, J. and Hamker, FH. (2007). "On the role of dopamine in cognitive vision", in Attention in Cognitive Systems, L. Paletta and E. Rome (Eds), Springer-Verlag LNAI 4840, pages 352-366.
  • Fix, J, Vitay, J. and Rougier, NP. (2006). "A computational model of spatial memory anticipation during visual search", in Anticipatory Behavior in Adaptive Learning Systems : From Brains to Individual and Social Behavior, M.V. Butz, O. Sigaud, G. Baldassarre and G. Pezzulo (Eds), Springer-Verlag LNAI 4520, pages 170-188.
  • Vitay, J., Rougier, NP. and Alexandre, F. (2005). "A distributed model of spatial visual attention", in Biomimetic Neural Learning for Intelligent Robotics, S. Wermter, G. Palm and M. Elshaw (Eds), Springer-Verlag LNAI 3575, pages 54-72.

International conferences

  • Vitay, J., Dinkelbach. HÜ. and Hamker, FH. (2014). "ANNarchy: Artificial Neural Networks architect", in Proceedings of BCCN'14 (Bernstein Conference on Computational Neuroscience), Tübingen (Germany), September 2014.
  • Vitay, J. and Hamker, FH. (2013). "A neurocomputational model of how VTA learns timing contingencies", at the ESF-FENS Conference on the Neurobiology of Action, Stresa (Italy), October 2013.
  • Vitay, J. and Hamker, FH. (2012). "Modulation of dopaminergic activity by the expectation of reward: a computational model", in Proceedings of BCCN'12 (Bernstein Conference on Computational Neuroscience), Munich (Germany), September 2012.
  • Vitay, J. and Hamker, FH. (2009). "A computational model of basal ganglia involved in the cognitive control of visual perception", in Proceedings of BCCN'09 (Bernstein Conference on Computational Neuroscience), Frankfurt (Germany), October 2009.
  • Vitay, J. and Hamker, FH. (2009). "Basal ganglia and memory retrieval during delayed match-to-sample and non-match-to-sample tasks", in Proceedings of CNS'09 (Annual Computational Neuroscience Meeting), Berlin (Germany), July 2009.
  • Vitay, J. and Hamker, FH. (2008). "Binding objects to cognition: A brain-like systems approach to the cognitive control of visual perception", in Proceedings of COGSYS'08 (International Conference on Cognitive Systems), Karlsruhe (Germany), April 2008.
  • Vitay, J. and Hamker, FH. (2008). "A computational model of how dopamine coordinates working memory and memory retrieval", in KogWiss'08 (Gesellschaft für Kognitionswissenschaft), Dresden (Germany), October 2008.
  • Vitay, J. and Hamker, FH. (2007). "Unspecific dopaminergic modulation of learning and working memory in perirhinal cortex", in Proceedings of ICVS'07 (International Conference on Computer Vision Systems), Bielefeld (Germany), March 2007.
  • Fix, J, Vitay, J. and Rougier, NP. (2006). "A computational model of spatial memory anticipation during visual search", in Proceedings of ABIALS'06 (Anticipatory Behavior in Adaptive Learning Systems), Roma (Italy), September 2006.
  • Vitay, J. (2006). "Emergence de fonctions par l'interaction de champs neuronaux", in Proceedings of NeuroComp (1ère Conférence Francophone sur les Neurosciences Computationnelles), Pont-à-Mousson (France), October 2006.
  • Vitay, J. (2005). "Towards teaching a robot to count objects", in Proceedings of Epirob'05 (Fifth International Workshop on Epigenetic Robotics), Nara (Japan), July 2005.
  • Vitay, J. and Rougier, NP. (2005). "Using neural dynamics to switch attention", in Proceedings of IJCNN'05 (International Joint Conference on Neural Networks), Montreal (Canada), July 2005.
  • Vitay, J., Rougier, NP. and Alexandre, F. (2004). "Reducing connectivity by using cortical modular bands", in Proceedings of ESANN'04 (European Symposium on Artificial Neural Networks), Brugges (Belgium), April 2004.


ANNarchy (Artificial Neural Networks architect)

ANNarchy (Artificial Neural Networks architect) is a parallel and hybrid simulator for distributed rate-coded or spiking neural networks. The core of the library is written in C++ and distributed using openMP or CUDA. It provides an interface in Python for the definition of the networks. It is released under the GNU GPL v2 or later.

Many modern neural simulators focus on the simulation of networks of spiking neurons on parallel hardware. Another important framework in computational neuroscience, rate-coded neural networks, is mostly difficult or impossible to implement using these simulators. We developed the ANNarchy (Artificial Neural Networks architect) neural simulator, which allows to easily define and simulate rate-coded and spiking networks, as well as combinations of both. The interface in Python has been designed to be close to the PyNN interface, while the definition of neuron and synapse models can be specified using an equation-oriented mathematical description similar to the Brian neural simulator. This information is used to generate C++ code that will efficiently perform the simulation on the chosen parallel hardware (multi-core system or graphical processing unit).


Machine Learning

Content: The course covers various fields of machine learning, including supervised learning, unsupervised learning and reinforcement learning. See the course's page : http://www.tu-chemnitz.de/informatik/KI/edu/ml

  1. Supervised Learning
    1. Linear Regression, gradient descent, Maximum-Likelihood
    2. Linear classification, Perceptron algorithm
    3. Multi-layer perceptron
    4. Support vector machines
    5. Deep Learning, Convolutional Neural Networks
    6. Unsupervised deep networks
    7. Recurrent neural networks
  2. Reinforcement Learning
    1. Formal definition of the RL-Problem
    2. Dynamic Programming, Monte-Carlo, Temporal Difference
    3. Eligibility Traces, Function approximation
    4. Deep Reinforcement Learning

Computer Vision

Content: The course covers different computer vision techniques, from simple image processing to top-down segmentation and feature matching. See the course's page : http://www.tu-chemnitz.de/informatik/KI/edu/biver

  1. Image formation
    1. Geometric primitives
    2. Color spaces
    3. Compression
  2. Image processing
    1. Histogram equalization
    2. Linear filtering
    3. Fourier transforms
    4. Pyramids and wavelets
  3. Geometric transformations
    1. Parametric transformations
    2. Mesh-based warping
  4. Feature detection and matching
    1. Feature detectors and descriptors
    2. Feature tracking
    3. Edge/Line detection
  5. Contours / Segmentation
    1. Active contours
    2. Split and merge
    3. Mean shift and mode finding
    4. Graph cuts
  6. Motion, optical flow
    1. Translational alignment
    2. Optical flow
    3. Layered motion
  7. Machine learning
    1. Bayes classifier
    2. Decision trees
    3. Boosting, random forest
    4. Haar Cascade
    5. Neural networks
  8. Feature-based alignment
    1. Panography
    2. Pose estimation
    3. Active appearance models

Introduction to Artificial Intelligence

Content: http://www.tu-chemnitz.de/informatik/KI/edu/ki.


StraNa, Room 1/348

Address: Fakultät für Informatik, Professur für Künstliche Intelligenz
Technische Universität Chemnitz
Straße der Nationen 62, D-09107 Chemnitz, Deutschland

Tel: (+49) 371 531 39468
Fax: (+49) 371 531 25739

Web: http://www.tu-chemnitz.de/~vitay
ORCID: orcid.org/0000-0001-5229-2349