Towards an open-source social middleware for humanoid robots

Sarabia, Miguel; Ros, Raquel; Demiris, Yiannis

doi:10.1109/humanoids.2011.6100883

Cited by 13 publications

(8 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It would also be interesting to investigate alternative hierarchical architectures (e.g., [26], [27]) for modelling more complex time-series. A further development on our classifier would be to incorporate actions, so that the iCub can re-grasp an object when uncertain about its class.…”

Section: Discussionmentioning

confidence: 99%

Online spatio-temporal Gaussian process experts with application to tactile classification

Soh

Demiris

2012

2012 IEEE/RSJ International Conference on Intelligent Robots and Systems

Self Cite

View full text Add to dashboard Cite

Abstract-In this work, we are primarily concerned with robotic systems that learn online and continuously from multivariate data-streams. Our first contribution is a new recursive kernel, which we have integrated into a sparse Gaussian Process to yield the Spatio-Temporal Online Recursive Kernel Gaussian Process (STORK-GP). This algorithm iteratively learns from time-series, providing both predictions and uncertainty estimates. Experiments on benchmarks demonstrate that our method achieves high accuracies relative to state-of-the-art methods. Second, we contribute an online tactile classifier which uses an array of STORK-GP experts. In contrast to existing work, our classifier is capable of learning new objects as they are presented, improving itself over time. We show that our approach yields results comparable to highly-optimised offline classification methods. Moreover, we conducted experiments with human subjects in a similar online setting with true-label feedback and present the insights gained.

show abstract

Section: Discussionmentioning

confidence: 99%

Online spatio-temporal Gaussian process experts with application to tactile classification

Soh

Demiris

2012

2012 IEEE/RSJ International Conference on Intelligent Robots and Systems

Self Cite

View full text Add to dashboard Cite

show abstract

“…The system described here is based on the integration of the latest HAMMER framework implementation [18] with a gaze controller which directs attention in order to maximise discrimination performance, while maintaining robustness to noise, and a contextual estimation of both end effector location and position of all potential targets.…”

Section: Action Recognition With Dynamic Allocation Of Attentionmentioning

confidence: 99%

Contextual action recognition and target localization with an active allocation of attention on a humanoid robot

et al. 2013

Self Cite

View full text Add to dashboard Cite

Abstract. Exploratory gaze movements are fundamental for gathering the most relevant information regarding the partner during social interactions. Inspired by the cognitive mechanisms underlying human social behaviour, we have designed and implemented a system for dynamic attention allocation which is able to actively control gaze movements during a visual action recognition task exploiting its own action execution predictions. Our humanoid robot is able, during the observation of a partner's reaching movement, to contextually estimate the goal position of the partner's hand and the location in space of the candidate targets. This is done while actively gazing around the environment, with the purpose of optimising the gathering of information relevant for the task. Experimental results on a simulated environment show that active gaze control, based on the internal simulation of actions, provides a relevant advantage with respect to other action perception approaches, both in terms of estimation precision and of time required to recognise and action. Moreover, our model reproduces and extends some experimental results on human attention during action perception.

show abstract

“…This framework has been implemented in real-dynamics robot simulators [17,14] and real robotic platforms [15,41,18]. Open source versions of the architecture have been freely released [60] with support for the NAO and iCub humanoids. Fig.…”

Section: Hierarchical Attentive Multiple Models For Execution and Recmentioning

confidence: 99%

Hierarchies for Embodied Action Perception

Ognibene

Lee

et al. 2013

Computational and Robotic Models of the Hierarchical Organization of Behavior

Self Cite

View full text Add to dashboard Cite

During social interactions, humans are capable of initiating and responding to rich and complex social actions despite having incomplete world knowledge as well as physical, perceptual and computational constraints. This capability relies on action perception mechanisms, which exploit regularities in observed goal-oriented behaviours to generate robust predictions, and reduce the workload of sensing systems. To achieve this essential capability, we argue that the following three factors are fundamental. Firstly, human knowledge is frequently hierarchically structured, both in the perceptual and execution domains. Secondly, human perception is an active process driven by current task requirements and context. This is particularly important when the perceptual input is complex (e.g. human motion) and the agent has to operate under embodiment constraints. Thirdly, learning is at the heart of action perception mechanisms, underlying the agent's ability to add new behaviours to its repertoire. Based on these factors, we review multiple instantiations of a hierarchically-organised biologically-inspired framework for embodied action perception, demonstrating its flexibility in addressing the rich computational contexts of action perception and learning in robotic platforms.

show abstract

Towards an open-source social middleware for humanoid robots

Cited by 13 publications

References 18 publications

Online spatio-temporal Gaussian process experts with application to tactile classification

Online spatio-temporal Gaussian process experts with application to tactile classification

Contextual action recognition and target localization with an active allocation of attention on a humanoid robot

Hierarchies for Embodied Action Perception

Contact Info

Product

Resources

About