Paul Verschure

Followers

142

Following

Co-authors

Public Views

Paul Verschure is Catalan Institute of Advanced Studies (ICREA) Research Professor, Director of the Center for Autonomous Systems and Neurorobotics at Universitat Pompeu Fabra and director of the neuro-engineering program at the Institute for Bioengineering of Catalunya, Barcelona Institute of Science and Technology. Paul directs the Synthetic Perceptive, Emotive and Cognitive Systems (SPECS) Laboratory (specs-lab.com). He is founder/CEO of Eodyne Systems S.L. (Eodyne.com), which is commercializing a novel science grounded neurorehabilitation technology. Paul is founder/Chairman of the Future Memory Foundation (futurememoryfoundation.org) which aims at supporting the development of new tools and paradigms for the conservation, presentation, and education of the history of the Holocaust and Nazi crimes. He received his MA and Ph.D. in Psychology, and Paul's scientific aim is to find a unified theory of mind and brain using synthetic methods and to apply it to the quality of life enhancing technologies. His theory of mind and brain, Distributed Adaptive Control, has been generalized to a range of brain structures and robotic systems and has laid the foundation for a novel neurorehabilitation approach called the Rehabilitation Gaming System (http://specs.upf.edu/research_in_neurorehabilitation). Paul explores new methods for the simulation, visualization, and exploration of complex data to support his DAC theory and advance clinical diagnostics and intervention in neuropathologies (brainx3.com). Complementary to his science, Paul has developed and deployed over 25 art installations (http://specs.upf.edu/installations). These include the biomimetic mixed reality space Ada experienced by over half a million visitors (2002) and more recently three virtual/augmented reality educational installations and applications for the Memorial Site Bergen-Belsen (2012 - ) which is now generalized to other sites across Europe.

less

Uploads

Papers by Paul Verschure

The EASEL Project: Towards Educational Human-Robot Symbiotic Interaction

Lecture Notes in Computer Science, 2016

This paper presents the EU EASEL project, which explores the potential impact and relevance of a ... more This paper presents the EU EASEL project, which explores the potential impact and relevance of a robot in educational settings. We present the project objectives and the theorectical background on which the project builds, briefly introduce the EASEL technological developments, and end with a summary of what we have learned from the evaluation studies carried out in the project so far.

Download

Modeling the Formation of Social Conventions in Multi-Agent Populations

arXiv (Cornell University), Feb 16, 2018

What is the role of real-time control and learning in the formation of social conventions? To ans... more What is the role of real-time control and learning in the formation of social conventions? To answer this question, we propose a computational model that matches human behavioral data in a social decision-making game that was analyzed both in discrete-time and continuous-time setups. Furthermore, unlike previous approaches, our model takes into account the role of sensorimotor control loops in embodied decision-making scenarios. For this purpose, we introduce the Control-based Reinforcement Learning (CRL) model. CRL is grounded in the Distributed Adaptive Control (DAC) theory of mind and brain, where low-level sensorimotor control is modulated through perceptual and behavioral learning in a layered structure. CRL follows these principles by implementing a feedback control loop handling the agent's reactive behaviors (pre-wired reflexes), along with an adaptive layer that uses reinforcement learning to maximize long-term reward. We test our model in a multi-agent game-theoretic task in which coordination must be achieved to find an optimal solution. We show that CRL is able to reach human-level performance on standard game-theoretic metrics such as efficiency in acquiring rewards and fairness in reward distribution.

Download

Towards a Synthetic Tutor Assistant: The EASEL Project and its Architecture

Lecture Notes in Computer Science, 2016

Robots are gradually but steadily being introduced in our daily lives. A paramount application is... more Robots are gradually but steadily being introduced in our daily lives. A paramount application is that of education, where robots can assume the role of a tutor, a peer or simply a tool to help learners in a specific knowledge domain. Such endeavor posits specific challenges: affective social behavior, proper modelling of the learner's progress, discrimination of the learner's utterances, expressions and mental states, which, in turn, require an integrated architecture combining perception, cognition and action. In this paper we present an attempt to improve the current state of robots in the educational domain by introducing the EASEL EU project. Specifically, we introduce the EASEL's unified robot architecture, an innovative Synthetic Tutor Assistant (STA) whose goal is to interactively guide learners in a science-based learning paradigm, allowing us to achieve such rich multimodal interactions.

Download

From motor to visually guided bimanual affordance learning

Adaptive Behavior, 2019

The mechanisms of how the brain orchestrates multi-limb joint action have yet to be elucidated an... more The mechanisms of how the brain orchestrates multi-limb joint action have yet to be elucidated and few computational sensorimotor (SM) learning approaches have dealt with the problem of acquiring bimanual affordances. We propose a series of bidirectional (forward/inverse) SM maps and its associated learning processes that generalize from uni- to bimanual interaction (and affordances) naturally, reinforcing the motor equivalence property. The SM maps range from a SM nature to a solely sensory one: full body control, delta SM control (through small action changes), delta sensory co-variation (how body-related perceptual cues covariate with object-related ones). We make several contributions on how these SM maps are learned: (1) Context and Behavior-Based Babbling: generalizing goal babbling to the interleaving of absolute and local goals including guidance of reflexive behaviors; (2) Event-Based Learning: learning steps are driven by visual, haptic events; and (3) Affordance Gradients...

Download

A model of grid cells based on a twisted torus topology

The grid cells of the rat medial entorhinal cortex (MEC) show an increased firing frequency when ... more The grid cells of the rat medial entorhinal cortex (MEC) show an increased firing frequency when the position of the animal correlates with multiple regions of the environment that are arranged in regular triangular grids. Here, we describe an artificial neural network based on a twisted torus topology, which allows for the generation of regular triangular grids. The association of the activity of pre-defined hippocampal place cells with entorhinal grid cells allows for a highly robust-to-noise calibration mechanism, suggesting a role for the hippocampal back-projections to the entorhinal cortex.

Download

A Cerebellar Neuroprosthetic System: Computational Architecture and in vivo Test

Emulating the input-output functions performed by a brain structure opens the possibility for dev... more Emulating the input-output functions performed by a brain structure opens the possibility for developing neuroprosthetic systems that replace damaged neuronal circuits. Here, we demonstrate the feasibility of this approach by replacing the cerebellar circuit responsible for the acquisition and extinction of motor memories. Specifically, we show that a rat can undergo acquisition, retention, and extinction of the eye-blink reflex even though the biological circuit responsible for this task has been chemically inactivated via anesthesia. This is achieved by first developing a computational model of the cerebellar microcircuit involved in the acquisition of conditioned reflexes and training it with synthetic data generated based on physiological recordings. Secondly, the cerebellar model is interfaced with the brain of an anesthetized rat, connecting the model's inputs and outputs to afferent and efferent cerebellar structures. As a result, we show that the anesthetized rat, equipped with our neuroprosthetic system, can be classically conditioned to the acquisition of an eye-blink response. However, non-stationarities in the recorded biological signals limit the performance of the cerebellar model. Thus, we introduce an updated cerebellar model and validate it with physiological recordings showing that learning becomes stable and reliable. The resulting system represents an important step toward replacing lost functions of the central nervous system via neuroprosthetics, obtained by integrating a synthetic circuit with the afferent and efferent pathways of a damaged brain region. These results also embody an early example of science-based medicine, where on the one hand the neuroprosthetic system directly validates a theory of cerebellar learning that informed the design of the system, and on the other one it takes a step toward the development of neuro-prostheses that could recover lost learning functions in animals and, in the longer term, humans.

Download

Interactive visuo-motor therapy system for stroke rehabilitation

Medical & Biological Engineering & Computing, 2007

We present a virtual reality (VR)-based motor neurorehabilitation system for stroke patients with... more We present a virtual reality (VR)-based motor neurorehabilitation system for stroke patients with upper limb paresis. It is based on two hypotheses: (1) observed actions correlated with self-generated or intended actions engage cortical motor observation, planning and execution areas (''mirror neurons''); (2) activation in damaged parts of motor cortex can be enhanced by viewing mirrored movements of non-paretic limbs. We postulate that our approach, applied during the acute post-stroke phase, facilitates motor re-learning and improves functional recovery. The patient controls a first-person view of virtual arms in tasks varying from simple (hitting objects) to complex (grasping and moving objects). The therapist adjusts weighting factors in the non-paretic limb to move the paretic virtual limb, thereby stimulating the mirror neuron system and optimizing patient motivation through graded task success. We present the system's neuroscientific background, technical details and preliminary results.

Download

The cerebellum chip: an analog VLSI implementation of a cerebellar model of classical conditioning

We present a biophysically constrained cerebellar model of classical conditioning, implemented us... more We present a biophysically constrained cerebellar model of classical conditioning, implemented using a neuromorphic analog VLSI (aVLSI) chip. Like its biological counterpart, our cerebellar model is able to control adaptive behavior by predicting the precise timing of events. Here we describe the functionality of the chip and present its learning performance, as evaluated in simulated conditioning experiments at the circuit level and in behavioral experiments using a mobile robot. We show that this aVLSI model supports the acquisition and extinction of adaptively timed conditioned responses under real-world conditions with ultra-low power consumption. * Jörg Kramer designed the cerebellum chip that was first tested at the 2002 Telluride Neuromorphic Engineering Workshop. Tragically, he died soon afterwards while hiking on Telescope Peak on

Download

iCub-HRI: A Software Framework for Complex Human–Robot Interaction Scenarios on the iCub Humanoid Robot

Frontiers in Robotics and AI, 2018

Generating complex, human-like behavior in a humanoid robot like the iCub requires the integratio... more Generating complex, human-like behavior in a humanoid robot like the iCub requires the integration of a wide range of open source components and a scalable cognitive architecture. Hence, we present the iCub-HRI library which provides convenience wrappers for components related to perception (object recognition, agent tracking, speech recognition, and touch detection), object manipulation (basic and complex motor actions), and social interaction (speech synthesis and joint attention) exposed as a C++ library with bindings for Java (allowing to use iCub-HRI within Matlab) and Python. In addition to previously integrated components, the library allows for simple extension to new components and rapid prototyping by adapting to changes in interfaces between components. We also provide a set of modules which make use of the library, such as a high-level knowledge acquisition module and an action recognition module. The proposed architecture has been successfully employed for a complex human-robot interaction scenario involving the acquisition of language capabilities, execution of goal-oriented behavior and expression of a verbal narrative of the robot's experience in the world. Accompanying this paper is a tutorial which allows a subset of this interaction to be reproduced. The architecture is aimed at researchers familiarizing themselves with the iCub ecosystem, as well as expert users, and we expect the library to be widely used in the iCub community.

Download

Neural correlates of visuomotor transformations using the Rehabilitation Gaming System (RGS): actual and imagined target catching

Klinische Neurophysiologie, 2012

A collision avoidance model based on the Lobula giant movement detector (LGMD) neuron of the locust

2004 IEEE International Joint Conference on Neural Networks (IEEE Cat. No.04CH37541), 2004

In insects we can find very complex and compact neural structures that are task specific. These n... more In insects we can find very complex and compact neural structures that are task specific. These neural structures allow them to perform complex tasks such as visual navigation, including obstacle avoidance, landing, self-stabilization, etc. Obstacle avoidance is fundamental for successful navigation, and it can be combined with more systems to make up more complex behaviors. In this paper we present a model for collision avoidance based on the Lobula Giant Movement Detector (LGMD) cell of the Locust. This is a wide-field visual neuron that responds to looming stimuli and that can trigger avoidance reactions whenever a rapidly approaching object is detected. Here we present result based on both an offline study of the model and its application to a flying robot.

Download

Learning Temporally Stable Representations from Natural Sounds: Temporal Stability as a General Objective Underlying Sensory Processing

Lecture Notes in Computer Science, 2007

In order to understand the general principles along which sensory processing is organized, severa... more In order to understand the general principles along which sensory processing is organized, several recent studies optimized particular coding objectives on natural inputs for different modalities. The homogeneity of neocortex indicates that a sensitive objective should be able to explain response properties of different sensory modalities. The temporal stability objective was successfully applied to somatosensory and visual processing. We investigate if this objective can also be applied to auditory processing and serves as a general optimization objective for sensory processing. In case of audition, this translates to a set of non-linear complex filters optimized for temporal stability on natural sounds. We show that following this approach we can develop filters that are localized in frequency and time and extract the frequency content of the sound wave. A subset of these filters respond invariant to the phase of the sound. A comparison of the tuning of these filters to the tuning of cat auditory nerves shows a close match. This suggests that temporal stability can be seen as a general objective describing somatosensory, visual and auditory processing.

Download

Robust Stimulus Encoding in Olfactory Processing: Hyperacuity and Efficient Signal Transmission

Lecture Notes in Computer Science, 2001

Download

Saliency Maps Operating on Stereo Images Detect Landmarks and Their Distance

Lecture Notes in Computer Science, 2002

We present a model that uses binocular visual input to detect landmarks and estimates their dista... more We present a model that uses binocular visual input to detect landmarks and estimates their distance based on disparity between the two images. Feature detectors provide input to saliency maps that find landmarks as combinations of features. Interactions between feature detectors for the left and right images and between the saliency maps enables corresponding landmarks to be found. We test the model in the real world and show that it reliably detects landmarks and estimates their distances.

Download

A Neural Model of the Fly Visual System Applied to Navigational Tasks

Lecture Notes in Computer Science, 2002

Download

Using a Mobile Robot to Study Locust Collision Avoidance Responses

International Journal of Neural Systems, 1999

The visual systems of insects perform complex processing using remarkably compact neural circuits... more The visual systems of insects perform complex processing using remarkably compact neural circuits, yet these circuits are often studied using simplified stimuli which fail to reveal their behaviour in more complex visual environments. We address this issue by testing models of these circuits in real-world visual environments using a mobile robot. In this paper we focus on the lobula giant movement detector (LGMD) system of the locust which responds selectively to objects which approach the animal on a collision course and is thought to trigger escape behaviours. We show that a neural network model of the LGMD system shares the preference for approaching objects and detects obstacles over a range of speeds. Our results highlight aspects of the basic response properties of the biological system which have important implications for the behavioural role of the LGMD.

Special Section on Virtual Rehabilitation

The dynamics of pattern recognition and categorization

Controlling the Movement of a Robot with a Brain-Computer Interface (BCI)

A closed-loop framework to investigate cerebellar learning in a decerebrated ferret preparation

Paul Verschure

Uploads

Papers by Paul Verschure

Log In