The MSNE program is organizing events and invited presentations open to students, staff, and members of the public. Members of the Elite Network of Bavaria are especially invited to this events. Attendance is free. Please register at the MSNE Events Mailinglist in order to receive upcoming event announcements by email. To register, please send an empty email to email@example.com
At some events: Within the first 10 minutes, 1-2 MSNE students are invited to present their research project results (short pitch presentation + poster). Students interested in presenting may approach firstname.lastname@example.org for the next available slots.
Prof. Dr. Shih-Chii Liu (ETH Zurich) Dec. 5th 2019 - 5 PM
Prof. Dr. Shih-Chii Liu
Institut für Neuroinformatik, University of Zürich /ETH Zürich
Title: Real-time Recognition with Neuromorphic Auditory Systems
Date, Time and Venue:Thursday, December 5th 2019, 17:00 , room N1135
Abstract: A fundamental organizing principle of brain computing enabling its amazing combination of intelligence, quick responsiveness, and low power consumption is its use of sparse spiking activity to drive computation. Recent progress in the development of higher-performance, more usable neuromorphic spike-event-based visual (DVS/ATIS/DAVIS) and auditory (AER-EAR/DAS) sensors along with versatile hardware such as FPGAs have stimulated exploration of real-time sensor processing for wearable and IoT platforms. These sensors enable "always-on" low-latency system-level response time at lower power than conventional sampled sensors. I will describe the circuits of a silicon cochlea auditory sensor that emulates the processing in biological cochleas, the event-driven deep networks that process the sensor data, and the real-time implementation of event-driven delta networks that emulate spiking networks on an FPGA platform with state-of-the-art power efficiency, latency, and throughput. I will demonstrate how we use these delta networks on a continuous spoken-digit speech recognition task.
Biography: Shih-Chii Liu received the B. S. degree in electrical engineering from Massachusetts Institute of Technology and the Ph.D. degree in the Computation and Neural Systems program from the California Institute of Technology in 1997. She worked at various companies including Gould American Microsystems, LSI Logic, and Rockwell International Research Labs before returning for her doctoral studies with Carver Mead. She is currently a Professor at the University of Zurich. Her research interests include low-power event-driven neuromorphic sensor and processor hardware design; and event-driven bio-inspired algorithms and deep neural networks, particularly for audio domains. Dr. Liu is past Chair of the IEEE CAS Sensory Systems and Neural Systems and Applications Technical Committees. She is current Chair of the IEEE Swiss CAS/ED Society and general co-chair of the 2020 IEEE Artificial Intelligence Circuits and Systems Conference (https://aicas2020.eu). She is also one of the lead organizers of the long-running Telluride Neuromorphic Cognition Engineering Workshop (http://tellurideneuromorphic.org). She co-directs the Sensors group (http://sensors.ini.uzh.ch) at the Institute of Neuroinformatics, University of Zurich and ETH Zurich.
The talk will be hosted by Prof. Werner Hemmert and Prof. Bernhard. U. Seeber
--> Download Flyer
Prof. Dr. Wulfram Gerstner (EPFL, Lausanne) Dec. 12th 2019 - 17:30
Prof. Dr. Wulfram Gerstner
Ecole polytechnique fédérale de Lausanne (EPFL)
Registration: We kindly ask to register quickly: https://wiki.tum.de/x/U4RJFQ
Date, Time and Venue: Th, Dec. 12th 2019, starting 17:30 at TUM ( Main Campus, Theresienstrasse 90, Building N1, room N1135)
Title: Eligibility traces and three-factor rules of synaptic plasticity
Abstract: Hebbian plasticity combines two factors: presynaptic activity must occur together with some postsynaptic variable (spikes, voltage deflection, calcium elevation ...). In three-factor learning rules the combination of the two Hebbian factors is not sufficient, but leaves a trace at the synapses (eligibility trace) which decays over a few seconds; only if a third factor (neuromodulator signal) is present, either simultaneously or within a short a delay, the actual change of the synapse via long-term plasticity is triggered. After a review of classic theories and recent evidence of plasticity traces from plasticity experiments in rodents, I will discuss two studies from my own lab: the first one is a modeling study of reward-based learning with spiking neurons using an actor-critic architecture; the second one is a joint theory-experimental study showing evidence for eligibility traces in human behavior and pupillometry. Extensions from reward-based learning to surprise-based learning will be indicated.
Biography: Wulfram Gerstner is Director of the Laboratory of Computational Neuroscience LCN at the EPFL. He studied physics at the universities of Tubingen and Munich and received a PhD from the Technical University of Munich. His research in computational neuroscience concentrates on models of spiking neurons and spike-timing dependent plasticity, on the problem of neuronal coding in single neurons and populations, as well as on the role of spatial representation for navigation of rat-like autonomous agents. He currently has a joint appointment at the School of Life Sciences and the School of Computer and Communications Sciences at the EPFL. He teaches courses for Physicists, Computer Scientists, Mathematicians, and Life Scientists. He is the recipient of the Valentino Braitenberg Award for Computational Neuroscience 2018 and a corresponding member of the Academy of Sciences and Literature Mainz (Germany).
The talk will be hosted by MSNE Students Jin Hwa Lee and Melanie Tschiersch.
In-House Visit at Texas Instruments (Jan 16th 2020)
MSNE Students interested in joining the event may register using the link provided in Newsletter. Please save the date! Tentative schedule: morning + early afternoon, agenda will be provided soon
Initiative by Prof. Bernhard Wolfrum & team.
Contact for administrative questions: email@example.com
Prof. Dr. Klaus Gramann (TU Berlin) - Mobile Brain/Body Imaging (MoBI) to image brains in action, Nov. 18th 2019, 17:30
Prof. Dr. Phil. Klaus Gramann
Institut für Psychologie und Arbeitswissenschaft Biopsychologie und Neuroergonomie https://www.bpn.tu-berlin.de/menue/team/klaus_gramann/
Title: Mobile Brain/Body Imaging (MoBI) to image brains in action
Monday, Nov 18th 2019, 17:30 at TUM, Theresienstrasse 90, in semianr room N1135
Abstract: Recent years have shown a remarkable shift in using established EEG technologies to leave the traditional lab environments and to record brain dynamics in actively behaving participants in complex technical setups and the real world. Imaging human brain dynamics usually requires stationary setups and immobile participants to avoid movement-related artifacts from distorting the signal of interest. Interaction with technical systems, however, often requires physical movement to interact with some form of interface to reach the desired system state. This movement itself provides kinesthetic feedback that contributes to and sometimes builds the very basis of the interaction. The brain dynamics underlying such physical interaction are hitherto unknown because of the restrictions of established brain imaging modalities. To overcome these restrictions, new mobile brain imaging methods can be employed. Here, I will give an overview of new technological developments in the field and applications that are now possible using mobile electroencephalography and Mobile Brain/Body Imaging (MoBI). The requirements and pros and cons of different approaches will be discussed and examples for recent interactive VR experiments are shown. In addition, challenges regarding recording hardware and analyses approaches often leading to difficulties in comparing the results with established laboratory EEG-parameters will be discussed.
Biography: Klaus Gramann received a pre-diploma in psychology from Justus Liebig University Giessen, Germany, and the Diploma and Ph.D. degrees in Psychology from RWTH Aachen, Germany in 1998 and 2002, respectively. He was an Assistant Professor with the Ludwig Maximilians University of Munich, Germany, and a Research Associate with the Swartz Center for Computational Neuroscience, University of California at San Diego. After working as a visiting professor at the National Chiao Tung University, Hsinchu, Taiwan and as a professor for cognitive psychology at the University of Osnabrueck, Germany, he became the chair of biopsychology and neuroergonomics with the Technical University of Berlin, Germany in 2012. Since 2017, he has also been a Professor with the School of Computer Science, University of Technology Sydney, Australia. His research covers the neural foundations of cognitive processes with a special focus on the brain dynamics of embodied cognitive processes. He is involved in the field of spatial cognition, visual attention, and the development of a mobile–brain imaging method to leverage the fundamental research results in applied neuroergonomics.
The talk will be hosted by Prof. Markus Ploner
Download Program Flyer
Prof. Dr. Dr. h.c. Robert Riener - Neurorehabilitation Robotics: Mechatronic Solutions for People with Movement Disorders (Nov. 7th 2019, 17:00)
Speaker: Prof. Dr.-Ing. Dr. med. h.c. Robert Riener
Sensory-Motor Systems Lab, IRIS, ETH Zurich and University of Zurich
Date and Venue: November 7th 2019, 17:00 (s.t.), room N1135
Title: Neurorehabilitation Robotics: Mechatronic Solutions for People with Movement Disorders
Abstract: Robots for the upper and lower limbs can be very useful to restore movement abilities in two ways. First, they can promote neurorehabilitation as training devices after neurological injuries such as spinal cord injury (SCI), traumatic brain injury and stroke. Second, they can be used as assistive devices to support patients or elders with gait impairments in daily life situations. However, current mechatronic solutions are still too bulky, too heavy, with too little battery power, and thus, too inconvenient to use. Furthermore, the sensory technologies and control strategies are still too primitive to allow the correct motion intention and to provide effective assist-as-needed support. These disadvantages result to unsatisfactory performance and discomfort. In this talk I will present current engineering solutions and future trends of stationary gait and arm training robots as well as wearable exoskeleton devices that can be used for training and assistance in daily life. I will also present the Cybathlon, which is a new kind of championship, where people with physical disabilities compete against each other at tasks of daily life, with the aid of robotic technologies. The next Cybathlon will take place in Zurich, on May 2nd and 3rd 2020.
Biography: Robert Riener studied Mechanical Engineering at TU München, Germany, and University of Maryland, USA. He received a Dr.-Ing. degree in Engineering from the TU München in 1997. After postdoctoral work from 1998-1999 at the Centro di Bioingegneria, Politecnico di Milano, he returned to TU München, where he completed his Habilitation in the field of Biomechatronics in 2003. In 2003 he became assistant professor at ETH Zurich and Spinal Cord Injury Center of the University Hospital Balgrist (“double-professorship”); since 2010 he has been full professor for Sensory-Motor Systems, ETH Zurich. Riener has published more than 400 peer-reviewed journal and conference articles, 20 books and book chapters and filed 24 patents. He has received 22 personal distinctions and awards including the IEEE TNSRE Best Paper Award 2010, and the euRobotics Technology Transfer Awards 2011 and 2012. Riener’s research focuses on the investigation of the sensory-motor interactions between humans and machines. Riener is the initiator and organizer of the Cybathlon, which was honored with the European Excellence Award and the Yahoo Sports Technology Award. In 2018 Riener obtained the honorary doctoral degree from the University of Basel.
The talk will be hosted by Prof. Gordon Cheng
Dr. Jessica Philipps-Silver - Auditory-Vestibulomotor Temporal Processing and Crossmodal Plasticity for Musical Rhythm in the Early Blind (Jun 27th 2019)
Laboratory of Integrative Neuroscience and Cognition
Georgetown University Medical Center
Title: Auditory-Vestibulomotor Temporal Processing and Crossmodal Plasticity for Musical Rhythm in the Early Blind
Abstract: The auditory dorsal stream (ADS) is a cortical brain network responsible for sensorimotor spatiotemporal processing. However, despite the important role of vestibular input when the head or body is moving through space, as well as the strong coupling between the vestibular and visual systems, very little is known about how vestibular information is integrated with auditory-motor inputs in the ADS, nor is it known to what extent this integration is affected by early visual deprivation. Using functional magnetic resonance imaging and motion capture technology we show that the ADS includes an extension to parietoinsular vestibular cortex (PIVC) and to subcortical regions including basal ganglia and vestibular cerebellum. This circuit is engaged after sensorimotor synchronization training, during beat recognition, and is preserved in the early blind. The strength of activation of PIVC in the early blind correlates with a measure of lifetime physical spatial activity, suggesting that experience with vestibular stimulation via physical spatial activities might compensate for any negative effects of early blindness, and thus reinforcing the beneficial effects of mobility training. Finally, rhythmic entrainment provides an effective tool for studying auditory-vestibulomotor integration and music appreciation, and for developing music-based interventions for early blind individuals.
Biography: Jessica Phillips-Silver, PhD, is a researcher in the Department of Neuroscience at Georgetown University Medical Center and served as adjunct professor in the Faculty of Music, where she developed Georgetown's first course on Music and the Brain. Jessica's research examines how 'feeling the beat' in music is a multisensory experience from infancy through adulthood, and she documented the first case of the musical disorder 'beat deafness'. She currently studies the musical processing and cortical plasticity in blindness with Prof. Josef Rauschecker at Georgetown. She is also interested in the development of musical rhythm and executive functions in Deaf and hearing children, and music and dance as a model of temporal prediction and cooperation in humans.
Date and Venue: Thursday, June 27th 2019 17:45 in room N1135
The talk ist hosted by ICS/ Prof. Gordon Cheng
Download Program Flyer
Prof. Dr. Alin Albu-Schäffer - Nonlinear elastic resonance modes for efficient robot and biological locomotion (Jun 26th 2019)
Prof. Dr.-Ing. Alin Albu-Schäffer
Deutsches Zentrum für Luft- und Raumfahrt (DLR)
Institut für Robotik und Mechatronik
Title: Nonlinear elastic resonance modes for efficient robot and biological locomotion
Abstract: Controlling motion at low energetic cost, both from mechanical and computational point of view, certainly constitutes one of the major locomotion challenges in biology and robotics. We attempt to demonstrate that robots can be designed and controlled to walk highly efficient by exploiting resonance body effects, increasing the performance compared to rigid body designs. To do so, however, legged robots need to achieve limit cycle motions of the highly coupled, non-linear body dynamics. This led us to the research of the still not very well understood theory of nonlinear system intrinsic modal oscillation control. I will present current theoretical and experimental results therewith. One of the striking results is that biomechanics, in particular kinematics, visco-elastic and inertial properties of biological limbs are such that coordinated resonant motions of multiple joints intrinsically emerges and is therefore easy to excite and sustain. This can be also achieved by careful design for robotic systems. Some of the basic robotics control functions we developed for locomotion strikingly resemble neural functionalities and structures. For example, Hebbian lerning, one of the most basic principles of synaptic plasticity, is mathematically equivalent to robotic controllers which adapt to previously unknown resonance properties of the body. Based on the robot control approach, we propose an equivalent neural model involving neural plasticity in the spine and the serotonergic loop in the brain-stem. This hypothesis is supported by numerous experimental evidences from neuroscience.
Biography: Alin Albu-Schäffer received his M.S. in electrical engineering from the Technical University of Timisoara, Romania in 1993 and his Ph.D. in automatic control from the Technical University of Munich in 2002. Since 2012 he is the head of the Institute of Robotics and Mechatronics at the German Aerospace Center (DLR), which he joined in 1995. Moreover, he is a professor at the Technical University of Munich, holding the Chair for "Sensor Based Robotic Systems and Intelligent Assistance Systems". His research interests range from robot design and control to robot intelligence and human neuroscience. He is an author of more than 250 peer reviewed journal and conference papers and received several awards, including the IEEE King-Sun Fu Best Paper Award of the Transactions on Robotics in 2012 and 2014;
Date and Venue: Jun 26th 2019 - 17:00 in room N1135
The Talk is hosted by Institute for Cognitive Systems (ICS, Prof. Gordon Cheng).
3rd MSNE Retreat in Brixlegg, Austria (May 30th - June 2nd 2019)
The 3rd MSNE Retreat will take place in the wonderful region of Tyrol (Austria), more precisely in Brixlegg. A four-day event (May 30th - June 2nd 2019), eintirely organized by MSNE students, for MSNE students and invited guests.
This year's invited talks / guest researchers are (in alphabetical order):
• Prof. Albert Compte / Insitut d'Investigations Biomèdiques August Pi y Sunyer
• Dr. Nora Heinzelmann, LMU, (Munich/Germany)
• Prof. Emily King, Bremen University, (Germany)
• Prof. Tobias Reichenbach from Imperial College London (UK)
The event is designed and organized by MSNE-Students. A few tickets are available for students from other related study programs. Interested students are encouraged to contact (firstname.lastname@example.org) latest by May 15th 2019.
Workshop on Intelligent Prosthetics and Brain-Computer Interfaces & Cybertum Hackathon (May 27th-29th 2019)
Cybertum Student Team Webpage (external service)
[Cybathlon] Excursion to REHAB Trade Fair, Karlsruhe (May 16th-18th 2019)
Teamleaders and supporters of MSNE/TUM teams for the 2020 ETH Cybathlon Challange may join a student delegation visiting the REHAB (Rehabilitation | Therapy | Care | Inclusion) trade fair in Karlsruhe. Contact by Zied Tayeb (ICS).
Excursion to Brainlab AG (May 16th 2019 - 13:00)
MSNE Students interested in joining the event may register until May 10th 2019 using the link provided in the Newsletter.
Dr. Ganesh Gowrishankar - Bad performance in sports is contagious:... (Feb 21st 2019)
Dr. Ganesh Gowrishankar, Laboratoire d'Informatique, de Robotique et de Microelectronique de Montpellier (LIRMM)
Title: Bad performance in sports is contagious: prediction error induced motor contagions in human behaviors
Abstract: Motor contagions refer to implicit effects on one’s actions induced by observed actions. Motor contagions are believed to be induced simply by action observation and cause an observer’s action to become similar to the action observed. However, in our recent work, we discovered a new motor contagion that is induced only when the observation is accompanied by prediction errors -differences between actions one observes and those he/she predicts or expects. In two experiments with professional sportsmen, we show that this contagion is distinct and arguably more dominant than contagions induced by action observation, and can lead to deterioration in performance of professionals after observing a novice. In this talk, I will give a brief summary of my work in “human centric robotics”- in which I utilize parallel research in robotics, motor neuroscience and cognitive neuroscience to improving machines that interact with humans. I will then talk in more detail about our cognitive neuroscience experiments investigating prediction error induced motor contagions, and discuss how this contagion may be the missing link between mechanisms investigated in action observation and action production by humans.
Biography: Gowrishankar Ganesh received his Bachelor of Engineering (first-class, Hons.) degree from the Delhi College of Engineering, India, in 2002 and his Master of Engineering from the National University of Singapore, in 2005, both in Mechanical Engineering. He received his Ph.D. in Bioengineering from Imperial College London, U.K., in 2010. He worked as an Intern Researcher with the Computational Neuroscience Laboratories, Advanced Telecommunication Research (ATR), Kyoto, Japan, from 2004 and through his PhD. Following his PhD he worked at the National Institute of Information and Communications Technology as a Specialist Researcher till December 2013. Since January 2014, he is a Senior Researcher at the Centre National de la Recherché Scientifique (CNRS), and is currently located at Le Laboratoire d'Informatique, de Robotique et de Microélectronique de Montpellier (LIRMM) in Montpellier. He is a visiting researcher at the University of Tokyo, National Institute of Advanced Industrial Science and Technology (AIST) in Tsukuba, and ATR in Kyoto. His research interests include robot control, human sensori-motor control and learning, cognitive neuroscience and robot-human interactions.
During the first 10 minutes, two MSNE students will present their research projects (part of their Research Excellence Certificate):
Suleman Zaidi - "Linguistic Analysis with Stereo EEG Data for functional Aphasia Location"
Francisco Zurita - "Influence of geometry on signal propagation in heartmuscle cell networks"
Time and Venue:
February 21st 2019 - 4:30 p.m., Theresienstarsse 90, Munich, room N1135
Talk is co-hosted by Prof. David Franklin (Neuromuscular Diagnostics) and Prof Gordon Cheng (Cognitive Systems)
Neuroengineering Matinee (Jan 16th 2019)
Srinivas Turaga, PhD - Connecting the structure and function of neural circuits - (Jan 16th 2019)
Srini Turaga, HHMI Janelia Research Campus
Connecting the structure and function of neural circuits
In this talk, I will describe how we developed deep learning based computational tools to solve two problems in neuroscience: inferring the activity of a neural network from measurements of its structural connectivity, and inferring the connectivity of a network of neurons from measurements and perturbation of neural activity.
1. Can we infer neural connectivity from noisy measurement and perturbation of neural activity? Population neural activity measurement by calcium imaging can be combined with cellular resolution optogenetic activity perturbations to enable the mapping of neural connectivity in vivo. This requires accurate inference of perturbed and unperturbed neural activity from calcium imaging measurements, which are noisy and indirect. We built on recent advances in variational autoencoderes to develop a new fully Bayesian approach to jointly inferring spiking activity and neural connectivity from in vivo all-optical perturbation experiments. Our model produces excellent spike inferences at 20K times real-time, and predicts connectivity for mouse primary visual cortex which is consistent with known measurements.
2. Are measurements of the structural connectivity of a biological neural network sufficient to predict its function? We constructed a simplified model of the first two stages of the fruit fly visual system, the lamina and medulla. The result is a deep hexagonal lattice convolutional neural network which discovered well-known orientation and direction selectivity properties in T4 neurons and their inputs. Our work demonstrates how knowledge of precise neural connectivity, combined with knowledge of the function of the circuit, can enable in silico predictions of the functional properties of individual neurons in a circuit, leading to an understanding of circuit function from structure.
Srini Turaga is a group leader at the Janelia Research Campus of the Howard Hughes Medical Institute. He was previously a postdoctoral fellow at the Gatsby Unit at University College London, following a PhD from MIT in 2009. His research interests include machine learning and computational neuroscience, with a special focus on connectomics, variational auto-encoders, and deep learning.
Time and Venue:
January 16th 2019 - 10.45 a.m., Arcisstrasse 21, Munich, room 5170 -> Vorhoelzer Forum
Talk is hosted by CNE (Prof. Jakob Macke).
Prof. Patrick van der Smagt - Latent optimal control (Jan 16th 2019)
Control of multidimensional systems typically relies on accurately engineered models. Breaking this require-ment is problematic with neural networks, as their Gaussian data assumptions typically do not hold. In my talk, I will demonstrate how this problem can be efficiently solved by combining latent variable models with specific type of optimal control. The theory is demonstrated on various simulated closed-loop control systems as well as on real hardware.
Patrick van der Smagt is director of the open-source Volkswagen Group AI Research Lab in Munich’s Volkswagen Data Lab, focusing on probabilistic deep learning for time series modelling, optimal control, rein-forcement learning, robotics, and quantum machine learning. He previously directed a lab as professor for machine learning and biomimetic robotics at the Technical University of Munich while leading the machine learning group at the research institute fortiss, and before founded and headed the Assistive Robotics and Bion-ics Lab at the DLR Oberpfaffenhofen. Quite a bit earlier, he did his PhD and MSc at Amsterdam’s universities. Besides publishing numerous papers and patents on machine learning, robotics, and motor control, he has won a number of awards, including the 2013 Helmholtz-Association Erwin Schrödinger Award, the 2014 King-Sun Fu Memorial Award, the 2013 Harvard Medical School/MGH Martin Research Prize, and best-paper awards at machine learning and robotics conferences and journals. He is founding chairman of a non-for-profit organisa-tion for Assistive Robotics for tetraplegics and co-founder of various tech companies.
Time and Venue:
January 16th 2019 - 10.00 a.m., Arcisstrasse 21, Munich, room 5170 -> Vorhoelzer Forum
Talk is hosted by CNE (Prof. Jakob Macke).
Dr. Josef Ladenbauer - Statistical inference for mechanistic models of neural populations based on spike-train data (December 11th 2018)
Abstract: Multi-neuronal spike-train data recorded in vivo typically exhibit rich dynamics as well as considerable variability across cells and repetitions of identical experimental conditions (trials). The interpretation of such data often relies on abstract statistical models that allow for principled parameter estimation and model selection; however, the interpretive power of these models is limited by the low extent to which prior biophysical constraints are incorporated. In contrast, mechanistic models are useful to interpret neurocircuit dynamics, but are rarely quantitatively matched to experimental data due to methodological challenges. In my talk I will present analytical, likelihood-based methods to efficiently fit spiking population models to single-trial spike trains. I will first focus on coupled stochastic integrate-and-fire neurons, for which we statistically infer the mean and variance of hidden inputs, neuronal adaptation properties and synaptic connectivity. Then, to infer the low-dimensional collective dynamics I will consider a doubly-stochastic model that accounts for fast independent and slower shared input fluctuations. We reconstruct the shared variations, classify their dynamics, obtain precise spike rate estimates, and quantify how individual neurons contribute to the population activity, all from a single trial. Extensive evaluations based on simulated data, and validations using ground truth recordings in vitro and in vivo demonstrate that our methods efficiently yield accurate results and outperform classical approaches. Altogether, these tools enable a quantitative, mechanistic interpretation of recorded neuronal population activity.
Time and Venue:
Tuesday, December 11th 2018, 16h00 Theresienstr. 90, 80333 Munich, room N1135
Talk is hosted by CNE (Prof. Jakob Macke)
Prof. Ryad B. Benosman - What is Neuromorphic Event-based Computer Vision? Sensors, Theory and Applications (November 12th 2018)
Abstract: In this presentation, I will introduce neuromorphic, event-based approaches for image sensing and processing. State-of-the-art image sensors suffer from severe limitations imposed by their very principle of operation. These sensors acquire the visual information as a series of “snapshots” recorded at discrete point in time, hence time-quantized at a predetermined frame rate, resulting in limited temporal resolution, low dynamic range and a high degree of redundancy in the acquired data. Nature suggests a different approach: Biological vision systems are driven and controlled by events happening within the scene in view, and not – like conventional image sensors – by artificially created timing and control signals that have no relation to the source of the visual information. Translating the frameless paradigm of biological vision to artificial imaging systems implies that control over the acquisition of visual information is no longer imposed externally on an array of pixels but rather the decision making is transferred to each individual pixel, which handles its own information individually. We will introduce the fundamentals underlying such bio-inspired, event-based image sensing and processing approaches, and explore their strengths and weaknesses. I will show that bio-inspired vision systems have the potential to wipe out conventional, frame-based vision acquisition and processing systems and to establish new benchmarks in terms of data compression, dynamic range, temporal resolution and power efficiency in applications such as 3D vision, object tracking, motor control, visual feedback loops, and machine learning in real-time at several hundreds kHz.
Ryad B. Benosman is a full Professor at both the University of Pittsburgh Medical Center, Carnegie Mellon University and Sorbonne Universitas where he does research at the intersection of robotics, computer vision and neuroscience. Specifically, he investigates the use of standard and neuromorphic cameras to enable au-tonomous, agile robotics, brain-machine interfaces focusing on retina prosthetics, optogenetics stimulation and recently visual cortex stimulation. Ryad did his PhD in robotics and computer vision at University of Pierre and Marie Curie after studying pure and applied mathematics. He is a pioneer and a leading researcher in the field of event based neuromorphic computer vision. His lab developed the ATIS neuromorphic camera. He has authored more than 200 papers, 60 of which are considered to provide the foundations of neuromorphic computer vision. He also founded sev-eral companies such as Prophesse (formerly Chronocam) the leading company in event-based vision, Pixium Vision (retina prosthetics), Chronolife (eHealth) and more recently Brainiac (neural processor computer).
Time and Venue:
Monday, November 12th 2018, 17h00 Theresienstr. 90, 80333 Munich, room N1135
Talk is hosted by ICS (Prof. Gordon Cheng)
Prof. Josef Rauschecker - Internal models of the brain in speech and music (June 26th 2018)
At first glance, the monkey brain looks like a smaller version of the human brain. Indeed, the anatomical and functional architecture of the cortical auditory system in monkeys is very similar to that of humans, with dual pathways segregated into a ventral and a dorsal processing stream. Yet, monkeys do not speak. Repeated attempts to pin this inability on one particular cause have failed. A closer look at the necessary components of language, according to Darwin, reveals that all of them got a significant boost during evolution from nonhuman to human primates. The vocal-articulatory system, in particular, has developed into the most sophisticated of all human sensorimotor systems with about a dozen effectors that, in combination with each other, result in an auditory communication system like no other. This sensorimotor network possesses all the ingredients of an internal model system that permits the emergence of sequence processing, as required for phonology and syn-tax in modern languages.
Josef P. Rauschecker studied at Technical University Munich (TUM) and Ludwig-Maximilians-University (LMU) Munich, Germany (Electrical Engineering and Medical Science) and at the Universities of Sussex (Ex-perimental Psychology and Artificial Intelligence) and Cambridge, England (Physiology). He received his Ph.D. (Dr.-Ing.) from TUM in 1980 for research performed at the Max Planck Institute (MPI) for Psychiatry in Munich and received his Habilitation (D.Sc.) in Neurophysiology from Eberhard-Karls-University Tübingen in 1985. After working as a junior staff scientist at the MPI for Biological Cybernetics from 1981-1989, he joined the Na-tional Institute of Mental Health (USA) as a Senior Investigator in the Laboratories of Neuropsychology and Neurophysiology in 1989. Since 1995, he has been a Professor of Physiology and Biophysics, Neurology, Psychology, and Neuroscience at Georgetown University, Washington, DC (USA), where he has also served on the university’s Executive Council, Steering Committee and Director of Cognitive Science. Josef Rauschecker is the director of the La-boratory of Integrative Neuroscience and Cognition (LINC) as well as of an international education and re-search Program in Cognitive and Computational Systems (PICCS).
Time and Venue:
Tuesday, June 26th 2018, 17h30 Theresienstr. 90, 80333 Munich, room N1135
Talk is hosted by ICS (Prof. Gordon Cheng).
Prof. Marc Spehr - Chemosensory Mechanisms of Conspecific Communication (Jan 10th 2018)
In most mammals, conspecific chemical communication controls complex behaviors. Information about individuality, social and reproductive status is conveyed by an elusive class of chemical cues – pheromones. The highly reproducible character of pheromone responses offers a unique opportunity to uncover the neuronal basis of genetically programmed behavior. Despite its fundamental significance, however, the basic chemosensory mechanisms of social communication remain largely unknown. To address these issues, my laboratory has developed a multi-faceted approach to uncover the mechanisms underlying mammalian pheromone sensing. My research, therefore, focuses on the molecular and cellular architecture of chemosensory communication in conspecific mammals – an innovative and interdisciplinary field of neurobiology. Combining molecular, biochemical, (electro)physiological, and live-cell imaging methods, as well as behavioral techniques in wildtype and mutant mouse models, my research challenges existing models of signal transduction in the olfactory system, analyzes the principle coding logic of pheromone detection, and, thus, sheds light on the neurophysiological basis of social behavior.
Marc Spehr is a Lichtenberg-Professor and head of the Chemosensation Laboratory at RWTH Aachen University in Aachen, Germany. He received his Diploma in biology as well as his Ph.D. (summa cum laude) from Ruhr-University Bochum, Germany. As a graduate student, Marc Spehr analyzed chemosensory signaling pathways in the mouse main and accessory olfactory systems. For his postdoctoral training he joined the group of Frank Zufall at University of Maryland School of Medicine in Baltimore, USA, where he investigated the role of the olfactory system in social recognition. In 2006, he was awarded an Emmy Noether grant by the German Research Counsel and returned to Ruhr-University Bochum as a principal investigator. As PI, his interests focused on the largely enigmatic function of pheromones in conspecific chemical communication. In wildtype and mutant mouse models, Marc Spehr addressed questions of both pheromone detection in the periphery and pheromone processing in the brain. In 2009, he was appointed a Lichtenberg-Professor of the Volkswagen Foundation at RWTH Aachen University where his laboratory continues to study the coding logic of pheromone detection and neurophysiological basis of social behavior.
Time and Venue:
Wednesday, January 10th 2018, 17h00 Theresienstr. 90, 80333 Munich, room N1135
Talk is hosted by the Neuroelectronics group (Prof. Wolfrum).
CoC Neuroengineering Networking Meeting - MSNE Symposium (19.4.2018)
Link to CoC Neuroengineering Event Page
A central part of the event is MSNE students' presentation on research projects and thesis in a conference/workshop style. Presenations are part of the optional MSNE Research Excellence Certificate most MSNE -Students are opting for.
Prof. Tom Francart - EEG in response to running speech: applications in diagnostics and noise suppression (Jan 15th 2018)
Tom Francart, born 1981 received the M.S.and Ph.D. degrees in engineering from the University of Leuven, Belgium, in 2004 and 2008, respectively. Since 2013 he is a research professor at the University of Leuven. His research interests include sound processing for auditory prostheses, binaural hearing and objective measures of hearing. His work follows a multidisciplinary approach that links electrical engineering with audiology and neuroscience. His current research focuses on the development of individualised and self-adapting sound processing for cochlear implant and hearing aid users.
The talk will be on EEG in response to running speech: applications in diagnostics and noise suppression.
Time and Venue:
Monday, January 15th 2018, 17:15 o'clock at Theresienstrasse 90, 80333 München, room N1135
The talk is hosted by Prof. Seeber (Audio Information Processing)
Dr. Mikhail A. Lebedev (Duke University) - Brain-Machine Interfaces for Restoration of Movements and Sensations (Feb 6th 2018)
Abstract: Brain-machine interfaces (BMIs) connect the nervous system to various external devices, with the goal of restoring and/or enhancing motor, sensory and cognitive functions, and neurorehabilitation. BMIs can be used by patients with different neurological conditions as assistive devices or by healthy individuals as tools for brain augmentation. Over the past half-century, BMIs have advanced significantly from the early ideas that sounded like science fiction to the modern high-tech implementations. In particular, invasive recordings using multichannel implants have enabled real-time control of artificial limbs by nonhuman primates and human subjects. Furthermore, neural prostheses can provide artificial sensory feedback, allowing users to perceive the movements of prosthetic limbs and their haptic interaction with external objects. Recently, neuroprosthetic approach was employed to build brain-nets that incorporate several brains exchanging information or performing cooperative tasks. Notwithstanding these achievements, even more spectacular developments are expected in the future.
Biography: Mikhail Lebedev, a Senior Research Scientist at Duke University, works in the fields of Neurorophysiology and Neuroprosthetics. He received a Master's degree in Physics from Moscow Institute of Physics and Technology (1986), and a PhD in Neurobiology from the University of Tennessee, Memphis (1995). His early research was on human sensorimotor integration. Since 1991, Lebedev works in the field of primate and rodent neurophysiology; he has studied neuronal mechanisms of cortical and basal ganglia circuits. Lebedev has investigated neuronal encoding of movements, somatic sensation, spatial attention and working memory. Since 2003, Lebedev works with Miguel Nicolelis; he supervises the primate laboratory at Duke University. The major focus of his current research is on BMIs, such as BMIs for reaching and grasping, BMIs that reproduce bipedal locomotion patterns, and sensorized BMIs that both extract motor command from the brain and deliver sensory information back to the brain. Lebedev is an editor of several journals, books and special issues.
Date, Time, and Venue: Tuesday, February 6th 2018, 16:00 o'clock at Theresienstrasse 90, 80333 München, room N1135
The talk is hosted by Prof. Cheng (Institute for Cognitive Systems)
Dr. Caroline Ling Li - Investigating into methods for multidimensional biomedical data analysis (13 January 2017)
ENB Elite Master Program Neuroengineering (MSNE)
This talk focuses on how the signal processing and data analysis methodologies can be applied into the biomedical signal analysis in the data fusion framework. In our analysis, brain activities are measured through Electroencephalography (EEG) signals which were obtained from intensive care unit in hospital. In order to identify the brain consciousness states of those patients, various signal processing and predictive analytics methods were used. In this talk, some advanced methods are presented which overcomes the weakness of traditional methods in both time and frequency domains for EEG analysis. Moreover, an online signal nature tracking method based on collaborative adaptive filter is also presented in order to monitoring the brain states in real time. We then discuss how these methodologies can be further extended as a general framework for studying human biological functions and performances, ranging from hand gesture recognition to sport sciences.
Dr. Caroline Ling Li has been a Lecturer in the school of Computing at the University of Kent since 2011. She is also the founding coordinator of Laboratory of Brain | Cognition | Computing (BC2 Lab) of the school responsible for coordinating multidisciplinary research between Computing, Sports and local NHS. In 2015, she organized the BIH’15 conference as the local chair, which is a gathering event for three of the biggest brain initiatives (Speakers include Prof Allan Jones, Prof Karlheinz Meier, Prof David Van Essen). Before she joined the University of Kent, she had six-year research experience at Imperial College London in signal processing with a focus of analyzing body sensor data (EEG, EMG, ECG, eAR-sensor, and etc.). She started her research study within the Department of Electrical and Electronic Engineering at Imperial and then worked as a research associate in the £6 million EPSRC “ESPRIT with Pervasive Sensing” project at the Department of Computing of Imperial College. She has been focused on developing advanced signal processing methods for understanding sensor data with biomedical applications such as EEG-based biomarker for brain diseases, EMG-controlled robotics, ECG pattern extraction, and human motion analysis.
Time and Venue
Friday, January 13th 2017, 14h15 , Karlstr. 45, 80333 Munich, Room 1025
Talk is hosted by the Professorship for Neuroscientific System Theory (Prof. Conradt).
Dr. Michael Pfeiffer - Deep Spiking Neural Networks – Low-Latency, Low-Compute Classifiers for Neuromorphic Platforms (20 December 2016)
ENB Elite Master Program Neuroengineering (MSNE)
Spiking neural networks (SNNs) originate from computational neuroscience, but in recent years there has been growing interest in using brain-inspired event-based computing for real-time pattern recognition. In my talk I will present novel approaches that merge ideas from SNNs and Deep Learning, the currently most successful machine learning paradigm for computer vision, speech recognition, and many real-world applications. Deep SNNs are particularly attractive for implementation in neuromorphic hardware platforms, which emulate the operation of spiking neurons in hardware, and achieve significant savings in power and latency over conventional models. New algorithmic insights allow us to reach accuracy levels that match traditional networks, while exploiting the advantages of SNNs. Deep SNNs exhibit a performance-latency tradeoff, which allows them to produce good first guesses very quickly, even before all neurons in the network are updated. I will show recent results that demonstrate how latency and computing costs in Deep SNNs can be reduced significantly, making them attractive models for fast and power-efficient information processing on power-constrained systems.
Dr. Michael Pfeiffer is a research engineer at Robert Bosch Corporate Research, where he investigates Cognitive Systems and Deep Learning. In 2010 he obtained his PhD in mathematics and computer science from Graz University of Technology, Austria, investigating machine learning methods as tools to understand computations in nervous systems. He then joined the Institute of Neuroinformatics at the University of Zurich and ETH Zurich as a postdoc, working on theories of neural computation and learning and neuromorphic computing. In 2012 he became group leader and program coordinator of the MSc in Neural Systems and Computation, an interdisciplinary specialized masters program combining systems neuroscience, theoretical neuroscience, neurotechnologies, and neuromorphic engineering. He has made substantial contributions towards understanding synaptic plasticity models such as STDP in the framework of machine learning. His work on deep and spiking neural networks has been influential for transferring recent breakthroughs from machine learning onto novel neuromorphic computing platforms
Time and Venue
Tuesday, December 20th 2016, 11h30 , Karlstr. 45, 80333 Munich, Room 2026
Talk is hosted by the Professorship for Neuroscientific System Theory (Prof. Conradt).
Prof. Maarten De Vos - Towards mobile and wearable brain monitoring
(25 November 2016)
ENB Elite Master Program Neuroengineering (MSNE)
All non-invasive technologies for the study of human brain activity suffer from the requirement that only artificial, movement-constrained behavior is allowed. However, by reducing “normal” behavior to a min-imum the ecological validity of the results can be limited. To overcome these limitations, we developed a truly mobile EEG system suitable for field recordings and natural situations which allows to decode single-trial brain responses in outdoor situations. We also demonstrated that signal quality of the mobile EEG system is equivalent to that of a standard lab amplifier in a traditional BCI experiment. Besides mobility and robustness with respect to motion, the critical issue before introducing EEG routinely in large studies became the electrode, as ideal EEG electrode would allow high quality and concealed recordings that can be conveniently attached to the head. We will demonstrate that a newly introduced cEEGrid electrode concept fulfills all these requirements and allows to monitor auditory attention relia-bly over long amounts of time.
Maarten De Vos is Associate Professor at the IBME, in the University of Oxford, following a Junior Professorship at the University of Oldenburg, Germany. He obtained his Ph.D. in electrical engineering from KU Leuven, Belgium, focusing on tensor-based decomposition methods. His academic work focuses on innovative biomedical monitoring and signal analysis, in particular the derivation of biosignatures of patient health from data acquired via wearable sensors and the incorporation of smart analytics into unobtrusive systems. He pioneering research in the field of mobile real-life brain-monitoring, which was awarded with several innovation prizes. His work on neonatal brain monitoring also achieved impact in patient care through the Neo-guard implementation project. After successful completion of the Biodesign faculty training at Stanford University, he started the Oxford Biodesign program.
Time and Venue
Friday, November 25th 2016, 15h00 at Karlstr. 45, 80333 Munich, Room 1025
Talk is hosted by the Professorship for Neuroscientific System Theory (Prof. Conradt).
Invitation Flyer (PDF)