A Dynamic Neural Field Approach to Natural and Efficient Human-Robot Collaboration

Erlhagen, Wolfram; Bicho, Estela

doi:10.1007/978-3-642-54593-1_13

Cited by 9 publications

(14 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This interpretation is also consistent with the reduced level of neural activity after pre-activation within a specific area4546. It is also consistent with computational assumptions about pre-activated neural populations taking advantage of populations at rest47. As N2cc was larger in UM than in OM, it is possible that cognitive control activity was more strongly modulated by previous activation/inactivation of the mechanisms involved than by the degree of interference that had to be managed.…”

Section: Discussionsupporting

confidence: 80%

Cognitive control activity is modulated by the magnitude of interference and pre-activation of monitoring mechanisms

Cespón

Galdo‐Álvarez

Dı́az

2016

Sci Rep

View full text Add to dashboard Cite

The Simon task is used to study interference from irrelevant spatial information. Interference is manifested by longer reaction times when the required response –based on non-spatial features- is spatially incompatible with stimulus position. Interference is greater when incompatible trials are preceded by compatible trials (compatible-incompatible sequence) than when they are preceded by incompatible trials (incompatible-incompatible sequence). However, the relationships between spatial attention, interference and cognitive control have not been investigated. In the present study, we distinguished three experimental conditions according to sequential effects: same mappings (SM, compatible-compatible/incompatible-incompatible sequences: low interference), opposite mappings (OM, compatible-incompatible/incompatible-compatible sequences: high interference) and unrelated mappings (UM, central-compatible/central-incompatible sequences: intermediate interference). The negativity central contralateral (N2cc, a correlate of prevention of spatial response tendencies) was larger in OM than in SM, indicating greater cognitive control for greater interference. Furthermore, N2cc was larger in UM than in SM/OM, indicating lower neural efficiency for suppressing spatial tendencies of the response after central trials. Attentional processes (negativity posterior contralateral) were also delayed in UM relative to SM/OM, suggesting attentional facilitation by similar sets of attentional shifts in successive trials. Overall, the present findings showed that cognitive control is modulated by the magnitude of interference and pre-activation of monitoring mechanisms.

show abstract

Section: Discussionsupporting

confidence: 80%

Cognitive control activity is modulated by the magnitude of interference and pre-activation of monitoring mechanisms

Cespón

Galdo‐Álvarez

Dı́az

2016

Sci Rep

View full text Add to dashboard Cite

show abstract

“…While the presented modeling work takes inspiration from neurophysiological and behavioral findings, it is also constrained by the specific needs of robotics applications. Our long-time goal is to develop autonomous robots able collaborate with humans in a natural that is human-like manner [9]. We have tested in the past a complex cognitive control architecture consisting of several coupled DNFs in a task in which the robot ARoS collaborates with different users in assembling a toy object from its parts [3].…”

Section: Discussionmentioning

confidence: 99%

Learning joint representations for order and timing of perceptual-motor sequences: A dynamic neural field approach

Wojtak

Ferreira

Erlhagen

et al. 2015

2015 International Joint Conference on Neural Networks (IJCNN)

Self Cite

View full text Add to dashboard Cite

Many of our everyday tasks require the control of the serial order and the timing of component actions. Using the dynamic neural field (DNF) framework, we address the learning of representations that support the performance of precisely time action sequences. In continuation of previous modeling work and robotics implementations, we ask specifically the question how feedback about executed actions might be used by the learning system to fine tune a joint memory representation of the ordinal and the temporal structure which has been initially acquired by observation. The perceptual memory is represented by a self-stabilized, multi-bump activity pattern of neurons encoding instances of a sensory event (e.g., color, position or pitch) which guides sequence learning. The strength of the population representation of each event is a function of elapsed time since sequence onset. We propose and test in simulations a simple learning rule that detects a mismatch between the expected and realized timing of events and adapts the activation strengths in order to compensate for the movement time needed to achieve the desired effect. The simulation results show that the effector-specific memory representation can be robustly recalled. We discuss the impact of the fast, activationbased learning that the DNF framework provides for robotics applications.

show abstract

“…A particular form of DNF first analyzed by Amari (1977), was used for modeling. In each layer i , the activity u i ( x , t ) at time t of a neuron at field location x is described in equation (1) (for mathematical details see Erlhagen & Bicho, 2014)…”

Section: Dynamical Neural Fields As a Theoretical Framework For Thementioning

confidence: 99%

“…In previous work, we have developed a cognitive control architecture for human-robot joint action that integrates action simulation, goal inference, error detection and complementary action selection (Bicho, Erlhagen, Louro, & Costa e Silva, 2011; Bicho, Erlhagen, Louro, Costa e Silva, Silva, & Hipólito, 2011), based on the neurocognitive mechanisms underlying human joint action (Bekkering et al, 2009). For the design and implementation, our group takes a neurodynamics approach based on the theoretical framework of Dynamic Neural Fields (DNFs) (Erlhagen & Bicho, 2006, 2014; Schöner, 2008). The robot is able to successfully collaborate with a human partner in joint tasks (e.g.…”

Section: Introductionmentioning

confidence: 99%

Combining intention and emotional state inference in a dynamic neural field architecture for human-robot joint action

et al. 2016

Self Cite

View full text Add to dashboard Cite

We report on our approach towards creating socially intelligent robots, which is heavily inspired by recent experimental findings about the neurocognitive mechanisms underlying action and emotion understanding in humans. Our approach uses neuro-dynamics as a theoretical language to model cognition, emotional states, decision making and action. The control architecture is formalized by a coupled system of dynamic neural fields representing a distributed network of local but connected neural populations. Different pools of neurons encode relevant information in the form of selfsustained activation patterns, which are triggered by input from connected populations and evolve continuously in time. The architecture implements a dynamic and flexible context-dependent mapping from observed hand and facial actions of the human onto adequate complementary behaviors of the robot that take into account the inferred goal and inferred emotional state of the co-actor. The dynamic control architecture was validated in multiple scenarios in which an anthropomorphic robot and a human operator assemble a toy object from its components. The scenarios focus on the robot's capacity to understand the human's actions, and emotional states, detect errors and adapt its behavior accordingly by adjusting its decisions and movements during the execution of the task.

show abstract

A Dynamic Neural Field Approach to Natural and Efficient Human-Robot Collaboration

Cited by 9 publications

References 44 publications

Cognitive control activity is modulated by the magnitude of interference and pre-activation of monitoring mechanisms

Cognitive control activity is modulated by the magnitude of interference and pre-activation of monitoring mechanisms

Learning joint representations for order and timing of perceptual-motor sequences: A dynamic neural field approach

Combining intention and emotional state inference in a dynamic neural field architecture for human-robot joint action

Contact Info

Product

Resources

About