Imperial College London

Professor Yiannis Demiris

Faculty of EngineeringDepartment of Electrical and Electronic Engineering

Professor of Human-Centred Robotics, Head of ISN
 
 
 
//

Contact

 

+44 (0)20 7594 6300y.demiris Website

 
 
//

Location

 

1014Electrical EngineeringSouth Kensington Campus

//

Summary

 

Publications

Publication Type
Year
to

191 results found

Carlson T, Demiris Y, 2012, Collaborative Control of a Robotic Wheelchair: Evaluation of Performance, Attention and Workload, IEEE Transactions on Systems, Man and Cybernetics, Part B: Cybernetics, Vol: 42, Pages: 876-888, ISSN: 1083-4419

Powered wheelchair users often struggle to drive safely and effectively and, in more critical cases, can only get around when accompanied by an assistant. To address these issues, we propose a collaborative control mechanism that assists users as and when they require help. The system uses a multiple-hypothesis method to predict the driver's intentions and, if necessary, adjusts the control signals to achieve the desired goal safely. The main emphasis of this paper is on a comprehensive evaluation, where we not only look at the system performance but also, perhaps more importantly, characterize the user performance in an experiment that combines eye tracking with a secondary task. Without assistance, participants experienced multiple collisions while driving around the predefined route. Conversely, when they were assisted by the collaborative controller, not only did they drive more safely but also they were able to pay less attention to their driving, resulting in a reduced cognitive workload. We discuss the importance of these results and their implications for other applications of shared control, such as brain-machine interfaces, where it could be used to compensate for both the low frequency and the low resolution of the user input.

Journal article

Lee K, Kim TK, Demiris Y, 2012, Learning Reusable Task Components using Hierarchical Activity Grammars with Uncertainties, St. Paul, Minnesota, USA, Publisher: IEEE, Pages: 1994-1999

We present a novel learning method using activity grammars capable of learning reusable task components from a reasonably small number of samples under noisy conditions. Our linguistic approach aims to extract the hierarchical structure of activities which can be recursively applied to help recognize unforeseen, more complicated tasks that share the same underlying structures. To achieve this goal, our method 1) actively searches for frequently occurring action symbols that are subset of input samples to effectively discover the hierarchy, and 2) explicitly takes into account the uncertainty values associated with input symbols due to the noise inherent in low-level detectors. In addition to experimenting with a synthetic dataset to systematically analyze the algorithm's performance, we apply our method in human-led imitation learning environment where a robot learns reusable components of the task from short demonstrations to correctly imitate more complicated, longer demonstrations of the same task category. The results suggest that under reasonable amount of noise, our method is capable to capture the reusable structures of tasks and generalize to cope with recursions.

Conference paper

Chatzis SP, Korkinof D, Demiris Y, 2012, The Kernel Pitman-Yor Process

In this work, we propose the kernel Pitman-Yor process (KPYP) fornonparametric clustering of data with general spatial or temporalinterdependencies. The KPYP is constructed by first introducing an infinitesequence of random locations. Then, based on the stick-breaking construction ofthe Pitman-Yor process, we define a predictor-dependent random probabilitymeasure by considering that the discount hyperparameters of theBeta-distributed random weights (stick variables) of the process are notuniform among the weights, but controlled by a kernel function expressing theproximity between the location assigned to each weight and the givenpredictors.

Journal article

Ognibene D, Demiris Y, 2012, Attentional shifts during action perception, Publisher: PION LTD, Pages: 1272-1272, ISSN: 0301-0066

Conference paper

Soh H, Demiris Y, 2012, Iterative Temporal Learning and Prediction with the Sparse Online Echo State Gaussian Process, International Joint Conference on Neural Networks, IJCNN, Publisher: IEEE, Pages: 1-8, ISSN: 2161-4393

In this work, we contribute the online echo state gaussian process (OESGP), a novel Bayesian-based online method that is capable of iteratively learning complex temporal dynamics and producing predictive distributions (instead of point predictions). Our method can be seen as a combination of the echo state network with a sparse approximation of Gaussian processes (GPs). Extensive experiments on the one-step prediction task on well-known benchmark problems show that OESGP produced statistically superior results to current online ESNs and state-of-the-art regression methods. In addition, we characterise the benefits (and drawbacks) associated with the considered online methods, specifically with regards to the trade-off between computational cost and accuracy. For a high-dimensional action recognition task, we demonstrate that OESGP produces high accuracies comparable to a recently published graphical model, while being fast enough for real-time interactive scenarios.

Conference paper

Ribes A, Cerquides J, Demiris Y, Lopez de Mantaras Ret al., 2012, Incremental Learning of an Optical Flow Model for Sensorimotor Anticipation in a Mobile Robot, IEEE International Conference on Development and Learning and Epigenetic Robotics (ICDL), Publisher: IEEE, ISSN: 2161-9484

Conference paper

Ribes A, Cerquides Bueno J, Demiris Y, Lopez de Mantaras Ret al., 2012, Context-GMM: Incremental Learning of Sparse Priors for Gaussian Mixture Regression, IEEE International Conference on Robotics and Biomimetics (ROBIO), Publisher: IEEE

Conference paper

Claassens J, Demiris Y, 2012, Exploiting Affordance Symmetries for Task Reproduction Planning, 12th IEEE-RAS International Conference on Humanoid Robots (Humanoids), Publisher: IEEE, Pages: 653-659, ISSN: 2164-0572

Conference paper

Chatzis SP, Demiris Y, 2012, The copula echo state network, Pattern Recognition, Vol: 45, Pages: 570-577, ISSN: 0031-3203

Echo state networks (ESNs) constitute a novel approach to recurrent neural network (RNN) training, with an RNN (the reservoir) being generated randomly, and only a readout being trained using a simple, computationally efficient algorithm. ESNs have greatly facilitated the practical application of RNNs, outperforming classical approaches on a number of benchmark tasks. This paper studies the formulation of a class of copula-based semiparametric models for sequential data modeling, characterized by nonparametric marginal distributions modeled by postulating suitable echo state networks, and parametric copula functions that help capture all the scale-free temporal dependence of the modeled processes. We provide a simple algorithm for the data-driven estimation of the marginal distribution and the copula parameters of our model under the maximum-likelihood framework. We exhibit the merits of our approach by considering a number of applications; as we show, our method offers a significant enhancement in the dynamical data modeling capabilities of ESNs, without significant compromises in the algorithm's computational efficiency.

Journal article

Chatzis SP, Korkinof D, Demiris Y, 2012, A spatially-constrained normalized Gamma process prior, Expert Systems with Applications, Vol: 39, Pages: 13019 - 13025-13019 - 13025, ISSN: 0957-4174

Journal article

Chatzis SP, Korkinof D, Demiris Y, 2012, A Quantum-Statistical Approach Toward Robot Learning by Demonstration, IEEE Transactions on Robotics, Vol: 28, Pages: 1371-1381-1371-1381, ISSN: 1941-0468

Statistical machine learning approaches have been at the epicenter of the ongoing research work in the field of robot learning by demonstration over the past few years. One of the most successful methodologies used for this purpose is a Gaussian mixture regression (GMR). In this paper, we propose an extension of GMR-based learning by demonstration models to incorporate concepts from the field of quantum mechanics. Indeed, conventional GMR models are formulated under the notion that all the observed data points can be assigned to a distinct number of model states (mixture components). In this paper, we reformulate GMR models, introducing some quantum states constructed by superposing conventional GMR states by means of linear combinations. The so-obtained quantum statistics-inspired mixture regression algorithm is subsequently applied to obtain a novel robot learning by demonstration methodology, offering a significantly increased quality of regenerated trajectories for computational costs comparable with currently state-of-the-art trajectory-based robot learning by demonstration approaches. We experimentally demonstrate the efficacy of the proposed approach.

Journal article

Ros R, Nalin M, Wood R, Baxter P, Looije R, Demiris Y, Belpaeme T, Giusti A, Pozzi Cet al., 2011, Child-robot interaction in the wild: Advice to the aspiring experimenter, Pages: 335-342

We present insights gleaned from a series of child-robot interaction experiments carried out in a hospital paediatric department. Our aim here is to share good practice in experimental design and lessons learned about the implementation of systems for social HRI with child users towards application in "the wild", rather than in tightly controlled and constrained laboratory environments: a trade-off between the structures imposed by experimental design and the desire for removal of such constraints that inhibit interaction depth, and hence engagement, requires a careful balance. © 2011 ACM.

Conference paper

Wu Y, Demiris Y, 2011, Learning Dynamical Representations of Tools for Tool-Use Recognition, International Conference on Robotics and Biomimetics (ROBIO), Publisher: IEEE, Pages: 2664-2669

We consider the problem of representing and recognising tools, a subset of objects that have special functionality and action patterns. Our proposed framework is based on the biological evidence of hierarchical representation of tools in the region of the human cortex that generates action semantics. It addresses the shortfalls of traditional learning models of object representation applied on tools. To showcase its merits, this framework is implemented as a hybrid model between the Hierarchical Attentive Multiple Models for Execution and Recognition of Actions Architecture (HAMMER) and Hidden Markov Model (HMM) to recognise and describe tools as dynamic patterns at symbolic level. The implemented model is tested and validated on two sets of experiments of 50 human demonstrations each on using 5 different tools. In the experiment with precise and accurate input data, the cross-validation statistics suggest very robust identification of the learned tools. In the experiment with unstructured environment, all errors can be explained systematically.

Conference paper

Chatzis SP, Korkinof D, Demiris Y, 2011, The One-Hidden Layer Non-parametric Bayesian Kernel Machine, IEEE International Conference on Tools with Artificial Intelligence (ICTAI)

In this paper, we present a nonparametric Bayesian approach towards one-hidden-layer feedforward neural net- works. Our approach is based on a random selection of the weights of the synapses between the input and the hidden layer neurons, and a Bayesian marginalization over the weights of the connections between the hidden layer neurons and the output neurons, giving rise to a kernel-based nonparametric Bayesian inference procedure for feedforward neural networks. Compared to existing approaches, our method presents a number of advan- tages, with the most significant being: (i) it offers a significant improvement in terms of the obtained generalization capabilities; (ii) being a nonparametric Bayesian learning approach, it entails inference instead of fitting to data, thus resolving the overfitting issues of non-Bayesian approaches; and (iii) it yields a full predictive posterior distribution, thus naturally providing a measure of uncertainty on the generated predictions (expressed by means of the variance of the predictive distribution), without the need of applying computationally intensive methods, e.g., bootstrap. We exhibit the merits of our approach by investigating its application to two difficult multimedia content classification applications: semantic characterization of audio scenes based on content, and yearly song classification, as well as a set of benchmark classification and regression tasks.

Conference paper

Claassens J, Demiris Y, 2011, Generalising human demonstration data by identifying affordance symmetries in object interaction trajectories., International Conference on Robotics and Intelligent Systems (IROS), Publisher: IEEE, Pages: 1980-1985

This paper concerns modelling human hand or tool trajectories when interacting with everyday objects. In these interactions symmetries may be exhibited in portions of the trajectories which can be used to identify task space redundancy. This paper presents a formal description of a set of these symmetries, which we term affordance symmetries, and a method to identify them in multiple demonstration recordings. The approach is robust to arbitrary motion before and after the symmetry artifact and relies only on recorded trajectory data. To illustrate the method's performance two examples are discussed involving two different types of symmetries. An simple illustration of the application of the concept in reproduction planning is also provided.

Conference paper

Kruijff-Korbayová I, Cuayáhuitl H, Kiefer B, Schröder M, Racioppa S, Cosi P, Sommavilla G, Tesser F, Sahli H, Athanasopoulos G, Wang W, Enescu V, Verhelst W, Cañamero L, Beck A, Hiolle A, Ros R, Demiris Yet al., 2011, A conversational system for multi-session child-robot interaction with several games, Germany, KI-2012: Poster and Demo Track, Pages: 135-139

Conference paper

Chatzis SP, Demiris Y, 2011, Echo state Gaussian process., IEEE Trans Neural Netw, Vol: 22, Pages: 1435-1445

Echo state networks (ESNs) constitute a novel approach to recurrent neural network (RNN) training, with an RNN (the reservoir) being generated randomly, and only a readout being trained using a simple computationally efficient algorithm. ESNs have greatly facilitated the practical application of RNNs, outperforming classical approaches on a number of benchmark tasks. In this paper, we introduce a novel Bayesian approach toward ESNs, the echo state Gaussian process (ESGP). The ESGP combines the merits of ESNs and Gaussian processes to provide a more robust alternative to conventional reservoir computing networks while also offering a measure of confidence on the generated predictions (in the form of a predictive distribution). We exhibit the merits of our approach in a number of applications, considering both benchmark datasets and real-world applications, where we show that our method offers a significant enhancement in the dynamical data modeling capabilities of ESNs. Additionally, we also show that our method is orders of magnitude more computationally efficient compared to existing Gaussian process-based methods for dynamical data modeling, without compromises in the obtained predictive performance.

Journal article

Chatzis SP, Demiris Y, 2011, Echo State Gaussian Process, IEEE Transactions on Neural Networks, Vol: 22, Pages: 1435-1445, ISSN: 1045-9227

Echo state networks (ESNs) constitute a novel approach to recurrent neural network (RNN) training, with an RNN (the reservoir) being generated randomly, and only a readout being trained using a simple computationally efficient algorithm. ESNs have greatly facilitated the practical application of RNNs, outperforming classical approaches on a number of benchmark tasks. In this paper, we introduce a novel Bayesian approach toward ESNs, the echo state Gaussian process (ESGP). The ESGP combines the merits of ESNs and Gaussian processes to provide a more robust alternative to conventional reservoir computing networks while also offering a measure of confidence on the generated predictions (in the form of a predictive distribution). We exhibit the merits of our approach in a number of applications, considering both benchmark datasets and real-world applications, where we show that our method offers a significant enhancement in the dynamical data modeling capabilities of ESNs. Additionally, we also show that our method is orders of magnitude more computationally efficient compared to existing Gaussian process-based methods for dynamical data modeling, without compromises in the obtained predictive performance.

Journal article

Lee K, Demiris Y, 2011, Towards incremental learning of task-dependent action sequences using probabilistic parsing, Frankfurt, Germany, International Conference on Development and Learning (ICDL), Publisher: IEEE, Pages: 1-6

We study an incremental process of learning where a set of generic basic actions are used to learn higher-level task-dependent action sequences. A task-dependent action sequence is learned by associating the goal given by a human demonstrator with the task-independent, general-purpose actions in the action repertoire. This process of contextualization is done using probabilistic parsing. We propose stochastic context-free grammars as the representational framework due to its robustness to noise, structural flexibility, and easiness on defining task-independent actions. We demonstrate our implementation on a real-world scenario using a humanoid robot and report implementation issues we had.

Conference paper

Soh H, Demiris Y, 2011, Evolving Policies for Multi-Reward Partially Observable Markov Decision Processes (MR-POMDPs), Genetic and Evolutionary Computation Conference (GECCO), Publisher: ACM, Pages: 713-720

Plans and decisions in many real-world scenarios are made under uncertainty and to satisfy multiple, possibly conflicting, objectives. In this work, we contribute the multi-reward partially-observable Markov decision process (MR-POMDP) as a general modelling framework. To solve MR-POMDPs, we present two hybrid (memetic) multi-objective evolutionary algorithms that generate non-dominated sets of policies (in the form of stochastic finite state controllers). Performance comparisons between the methods on multi-objective problems in robotics (with 2, 3 and 5 objectives), web-advertising (with 3, 4 and 5 objectives) and infectious disease control (with 3 objectives), revealed that memetic variants outperformed their original counterparts. We anticipate that the MR-POMDP along with multi-objective evolutionary solvers will prove useful in a variety of theoretical and real-world applications.

Conference paper

Soh H, Demiris Y, 2011, Involving Young Children in the Design of a Safe, Smart Paediatric Wheelchair, HRI Pioneers Workshop, Pages: 86-87

Independent mobility is crucial for a growing child and its loss can severely impact cognitive, emotional and social development. Unfortunately, powered wheelchair provision for young children has been difficult due to safety concerns. But powered mobility need not be unsafe. Risks can be reduced through the use of robotic technology (e.g., obstacle avoidance) and we present a prototype safe smart paediatric wheelchair: the Assistive Robot Transport for Youngsters (ARTY). A core aspect of our work is that we aim to bring ARTY to the field and we discuss the challenges faced when trying to involve children in the development/testing of medical technology. We discuss one preliminary experiment designed as a “Hide-and-Seek” game as a short case study.

Conference paper

Ros R, Demiris Y, Baroni I, Nalin Met al., 2011, Adapting Robot Behavior to User's Capabilities: A Dance Instruction Study, 6th ACM/IEEE International Conference on Human-Robot Interaction (HRI), Publisher: IEEE, Pages: 235-236, ISSN: 2167-2121

Conference paper

Soh H, Demiris Y, 2011, Multi-reward policies for medical applications, the 13th annual conference companion, Publisher: ACM Press

Conference paper

Soh H, Demiris Y, 2011, Multi-reward policies for medical applications: anthrax attacks and smart wheelchairs., Publisher: ACM, Pages: 471-478

Conference paper

Sarabia M, Ros R, Demiris Y, 2011, Towards an open-source social middleware for humanoid robots, International Conference on Humanoid Robotics, Publisher: IEEE, Pages: 670-675

Recent examples of robotics middleware including YARP, ROS, and NaoQi, have greatly enhanced the standardisation, interoperability and rapid development of robotics application software. In this paper, we present our research towards an open source middleware to support the development of social robotic applications. In the core of the ability of a robot to interact socially are algorithms to perceive the actions and intentions of a human user. We attempt to provide a computational layer to standardise these algorithms utilising a bioinspired computational architecture known as HAMMER (Hierarchical Attentive Multiple Models for Execution and Recognition) and demonstrate the deployment of such layer on two different humanoid platforms, the Nao and iCub robots. We use a dance interaction scenario to demonstrate the utility of the framework.

Conference paper

Takacs B, Demiris Y, 2010, Spectral clustering in multi-agent systems, Knowledge and Information Systems, Vol: 25, Pages: 607-622, ISSN: 0219-1377

We examine the application of spectral clustering for breaking up the behavior of a multi-agent system in space and time into smaller, independent elements. We propose clustering observations of individual entities in order to identify significant changes in the parameter space (like spatial position) and detect temporal alterations of behavior within the same framework. Available knowledge of important interactions (events) between entities is also considered. We describe a novel algorithm utilizing iterative subdivisions where clusters are pre-processed at each step to counter spatial scaling, rotation, replay speed, and varying sampling frequency. A method is presented to balance spatial and temporal segmentation based on the expected group size, and a validity measure is introduced to determine the optimal number of clusters. We demonstrate our results by analyzing the outcomes of computer games and compare our algorithm to K-means and traditional spectral clustering.

Journal article

Wu Y, Kuvinichkul P, Cheung PYK, Demiris Yet al., 2010, Towards Anthropomorphic Robot Thereminist, International Conference on Robotics and Biomimetics (ROBIO), Publisher: IEEE, Pages: 235-240

Theremin is an electronic musical instrument considered to be the most difficult to play which requires the player's hands to have high precision and stability as any position change within proximity of the instrument's antennae can make a difference to the pitch or volume. In a different direction to previous developments of Theremin playing robots, we propose a Humanoid Thereminist System that goes beyond using only one degree of freedom which will open up the possibility for robot to acquire more complex skills, such as aerial fingering and include musical expressions in playing the Theremin. The proposed system consists of two phases, namely calibration phase and playing phase which can be executed independently. During the playing phase, the System takes input from a MIDI file and performs path planning using a combination of minimum energy strategy in joint space and feedback error correction for next playing note. Three experiments have been conducted to evaluate the developed system quantitatively and qualitatively by playing a selection of music files. The experiments have demonstrated that the proposed system can effectively utilise multiple degrees of freedoms while maintaining minimum pitch error margins.

Conference paper

Wu Y, Demiris Y, 2010, Hierarchical Learning Approach for One-shot Action Imitation in Humanoid Robots, International Conference on Control, Automation, Robotics and Vision (ICARCV), Publisher: IEEE, Pages: 453-458

We consider the issue of segmenting an action in the learning phase into a logical set of smaller primitives in order to construct a generative model for imitation learning using a hierarchical approach. Our proposed framework, addressing the “how-to” question in imitation, is based on a one-shot imitation learning algorithm. It incorporates segmentation of a demonstrated template into a series of subactions and takes a hierarchical approach to generate the task action by using a finite state machine in a generative way. Two sets of experiments have been conducted to evaluate the performance of the framework, both statistically and in practice, through playing a tic-tac-toe game. The experiments demonstrate that the proposed framework can effectively improve the performance of the one-shot learning algorithm and reduce the size of primitive space, without compromising the learning quality.

Conference paper

Martins MF, Demiris Y, 2010, Learning multirobot joint action plans from simultaneous task execution demonstrations., International Conference on Autonomous Agents and Multiagent Systems, Publisher: ACM, Pages: 931-938

The central problem of designing intelligent robot systems which learn by demonstrations of desired behaviour has been largely studied within the field of robotics. Numerous architectures for action recognition and prediction of intent of a single teacher have been proposed. However, little work has been done addressing how a group of robots can learn by simultaneous demonstrations of multiple teachers.This paper contributes a novel approach for learning multirobot joint action plans from unlabelled data. The robots firstly learn the demonstrated sequence of individual actions using the HAMMER architecture. Subsequently, the group behaviour is segmented over time and space by applying a spatio-temporal clustering algorithm.The experimental results, in which humans teleoperated real robots during a search and rescue task deployment, successfully demonstrated the efficacy of combining action recognition at individual level with group behaviour segmentation, spotting the exact moment when robots must form coalitions to achieve the goal, thus yielding reasonable generation of multirobot joint action plans.

Conference paper

Wu Y, Demiris Y, 2010, Towards One Shot Learning by Imitation for Humanoid Robots, International Conference on Robotics and Automation (ICRA), Publisher: IEEE, Pages: 2889-2894, ISSN: 1050-4729

Teaching a robot to learn new knowledge is a repetitive and tedious process. In order to accelerate the process, we propose a novel template-based approach for robot arm movement imitation. This algorithm selects a previously observed path demonstrated by a human and generates a path in a novel situation based on pairwise mapping of invariant feature locations present in both the demonstrated and the new scenes using a combination of minimum distortion and minimum energy strategies. This One-Shot Learning algorithm is capable of not only mapping simple point-to-point paths but also adapting to more complex tasks such as those involving forced waypoints. As compared to traditional methodologies, our work require neither extensive training for generalisation nor expensive run-time computation for accuracy. This algorithm has been statistically validated using cross-validation of grasping experiments as well as tested for practical implementation on the iCub humanoid robot for playing the tic-tac-toe game.

Conference paper

This data is extracted from the Web of Science and reproduced under a licence from Thomson Reuters. You may not copy or re-distribute this data in whole or in part without the written consent of the Science business of Thomson Reuters.

Request URL: http://wlsprd.imperial.ac.uk:80/respub/WEB-INF/jsp/search-html.jsp Request URI: /respub/WEB-INF/jsp/search-html.jsp Query String: id=00333953&limit=30&person=true&page=4&respub-action=search.html