Spiking Deep Neural Networks

We have developed some of the first spiking implementations of deep neural networks for visual object recognition and sensory fusion. These networks allow to utilize the power and accuracy of deep learning models in real time, with short latencies and with much higher efficiency in terms of power requirements and necessary update steps. They are also ideal models for processing neuromorphic sensor inputs, and for implementations on special purpose hardware platforms, such as SpiNNaker.
Students:

• Danny Neil
• Jonathan Binas
• Iulia-Alexandra Lungu
• Bodo Rückauer

Collaborators:

• Shih-Chii Liu
• Tobi Delbruck
• Peter Diehl
• Steve Furber

Publications:

• P. O'Connor, D. Neil, S-C Liu, T. Delbruck, M. Pfeiffer, Real-Time Classification and Sensor Fusion with a Spiking Deep Belief Network. Frontiers in Neuromorphic Engineering 7(178), 2013. [link]
• E. Stromatias, D. Neil, F. Galluppi, M. Pfeiffer, S-C Liu, and S. Furber: Live Demonstration: Handwritten digit recognition using spiking Deep Belief Networks on SpiNNaker. IEEE Int. Symposium on Circuits and Systems (ISCAS), Lissabon, Portugal. 2015.
• E. Stromatias, D. Neil, F. Galluppi, M. Pfeiffer, S-C Liu, and S. Furber: Scalable Energy-Efficient, Low-Latency Implementations of Spiking Deep Belief Networks on SpiNNaker. Int. Joint Conference on Neural Networks (IJCNN), Killarney, Ireland. 2015.
• P. Diehl, D. Neil, J. Binas, M. Cook. S-C Liu, and M. Pfeiffer: Fast-Classifying, High-Accuracy Spiking Deep Networks Through Weight and Threshold Balancing. Int. Joint Conference on Neural Networks (IJCNN), Killarney, Ireland. 2015. [PDF]
• D. Neil, M. Pfeiffer, and S-C Liu.: Learning to be efficient: Learning to be Efficient: Algorithms for Training Low-Latency, Low-Compute Deep Spiking Neural Networks. ACM Symposium on Applied Computing, Pisa, Italy. 2016. [PDF]

Spike-based computation and learning in distributed neuromorphic systems

Together with Giacomo Indiveri I am leading this project, sponsored by the Swiss National Science Foundation (SNSF). The project investigates ways of configuring neuromorphic computing devices to enable functions such as learning, state-dependent computation, and probabilistic inference.
Students:

• Dora Sumislawska
• Jonathan Binas

Publications:

• J. Binas, U. Rutishauser, G. Indiveri, M. Pfeiffer, Learning and Stabilization of Winner-Take-All Dynamics Through Interacting Excitatory and Inhibitory Plasticity. Frontiers in Computational Neuroscience 8(68), 2014. [link]
• F. Galluppi, X. Lagorce, E. Stromatias, M. Pfeiffer, L. Plana, S. Furber, R. Benosman, A framework for plasticity implementation on the SpiNNaker neural architecture. Frontiers in Neuromorphic Engineering 8:429, 2014. [link]
• P. Diehl, D. Neil, J. Binas, M. Cook. S-C Liu, and M. Pfeiffer: Fast-Classifying, High-Accuracy Spiking Deep Networks Through Weight and Threshold Balancing. Int. Joint Conference on Neural Networks (IJCNN), Killarney, Ireland. 2015.
• J. Binas, G. Indiveri, and M. Pfeiffer: Local Structure Helps Learning Optimized Automata in Recurrent Neural Networks. Int. Joint Conference on Neural Networks (IJCNN), Killarney, Ireland. 2015.
• J. Binas, G. Indiveri, M. Pfeiffer, Spiking Analog VLSI Neuron Assemblies as Constraint Satisfaction Problem Solvers. IEEE Int. Symposium on Circuits and Systems (ISCAS), Montreal, Canada. 2016. [PDF]
• D. Sumislawska, N. Qiao, M. Pfeiffer, G. Indiveri, Wide dynamic range weights and biologically realistic synaptic dynamics for spike-based learning circuits. IEEE Int. Symposium on Circuits and Systems (ISCAS), Montreal, Canada. 2016.
• J. Binas, D. Neil, G. Indiveri, S-C Liu, M. Pfeiffer, Precise deep neural network computation on imprecise low-power analog hardware. arXiv 1606.07786. 2016. [arXiv]

A computational analysis of primate-specific features of corticogenesis

Marion Betizeau has been awarded a Transition Postdoc Fellowship by SystemsX.ch to investigate new data analysis methods based on Hidden Markov Trees to analyze cell lineages in primate cortical development.
Postdoc:

• Marion Betizeau

Collaborators:

• Rodney Douglas
• Gabriela Michel
• Saray Soldado Magraner
• Colette Dehay
• Henry Kennedy

Publications:

• M. Pfeiffer, M. Betizeau, J. Waltispurger, S. Pfister, RJ Douglas, H. Kennedy, C. Dehay, Unsupervised lineage-based characterization of primate precursors reveals high proliferative and morphological diversity in the OSVZ. Journal of Comparative Neurology, in press, 2015.

Last changed: 03 December 2015