Publications
Results
- Showing results for:
- Reset all filters
Search results
-
Conference paperKanajar P, Caldwell DG, Kormushev P, 2017,
Climbing over large obstacles with a humanoid robot via multi-contact motion planning
, IEEE RO-MAN 2017: 26th IEEE International Symposium on Robot and Human Interactive Communication, Publisher: IEEE, Pages: 1202-1209Incremental progress in humanoid robot locomotion over the years has achieved important capabilities such as navigation over flat or uneven terrain, stepping over small obstacles and climbing stairs. However, the locomotion research has mostly been limited to using only bipedal gait and only foot contacts with the environment, using the upper body for balancing without considering additional external contacts. As a result, challenging locomotion tasks like climbing over large obstacles relative to the size of the robot have remained unsolved. In this paper, we address this class of open problems with an approach based on multi-body contact motion planning guided through physical human demonstrations. Our goal is to make the humanoid locomotion problem more tractable by taking advantage of objects in the surrounding environment instead of avoiding them. We propose a multi-contact motion planning algorithm for humanoid robot locomotion which exploits the whole-body motion and multi-body contacts including both the upper and lower body limbs. The proposed motion planning algorithm is applied to a challenging task of climbing over a large obstacle. We demonstrate successful execution of the climbing task in simulation using our multi-contact motion planning algorithm initialized via a transfer from real-world human demonstrations of the task and further optimized.
-
Conference paperZhang F, Cully A, Demiris YIANNIS, 2017,
Personalized Robot-assisted Dressing using User Modeling in Latent Spaces
, 2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Publisher: IEEE, ISSN: 2153-0866Robots have the potential to provide tremendous support to disabled and elderly people in their everyday tasks, such as dressing. Many recent studies on robotic dressing assistance usually view dressing as a trajectory planning problem. However, the user movements during the dressing process are rarely taken into account, which often leads to the failures of the planned trajectory and may put the user at risk. The main difficulty of taking user movements into account is caused by severe occlusions created by the robot, the user, and the clothes during the dressing process, which prevent vision sensors from accurately detecting the postures of the user in real time. In this paper, we address this problem by introducing an approach that allows the robot to automatically adapt its motion according to the force applied on the robot's gripper caused by user movements. There are two main contributions introduced in this paper: 1) the use of a hierarchical multi-task control strategy to automatically adapt the robot motion and minimize the force applied between the user and the robot caused by user movements; 2) the online update of the dressing trajectory based on the user movement limitations modeled with the Gaussian Process Latent Variable Model in a latent space, and the density information extracted from such latent space. The combination of these two contributions leads to a personalized dressing assistance that can cope with unpredicted user movements during the dressing while constantly minimizing the force that the robot may apply on the user. The experimental results demonstrate that the proposed method allows the Baxter humanoid robot to provide personalized dressing assistance for human users with simulated upper-body impairments.
-
Conference paperRakicevic N, Kormushev P, 2017,
Efficient Robot Task Learning and Transfer via Informed Search in Movement Parameter Space
, Workshop on Acting and Interacting in the Real World: Challenges in Robot Learning, 31st Conference on Neural Information Processing Systems (NIPS 2017) -
Conference paperTavakoli A, Pardo F, Kormushev P, 2017,
Action Branching Architectures for Deep Reinforcement Learning
, Deep Reinforcement Learning Symposium, 31st Conference on Neural Information Processing Systems (NIPS 2017) -
Conference paperSalimbeni H, Deisenroth M, 2017,
Doubly stochastic variational inference for deep Gaussian processes
, NIPS 2017, Publisher: Advances in Neural Information Processing Systems (NIPS), Pages: 4589-4600, ISSN: 1049-5258Gaussian processes (GPs) are a good choice for function approximation as theyare flexible, robust to over-fitting, and provide well-calibrated predictiveuncertainty. Deep Gaussian processes (DGPs) are multi-layer generalisations ofGPs, but inference in these models has proved challenging. Existing approachesto inference in DGP models assume approximate posteriors that forceindependence between the layers, and do not work well in practice. We present adoubly stochastic variational inference algorithm, which does not forceindependence between layers. With our method of inference we demonstrate that aDGP model can be used effectively on data ranging in size from hundreds to abillion points. We provide strong empirical evidence that our inference schemefor DGPs works well in practice in both classification and regression.
-
Conference paperEleftheriadis S, Nicholson TFW, Deisenroth MP, et al., 2017,
Identification of Gaussian Process State Space Models
, Advances in Neural Information Processing Systems (NIPS) 2017, Publisher: Neural Information Processing Systems Foundation, Inc., Pages: 5310-5320, ISSN: 1049-5258The Gaussian process state space model (GPSSM) is a non-linear dynamicalsystem, where unknown transition and/or measurement mappings are described byGPs. Most research in GPSSMs has focussed on the state estimation problem.However, the key challenge in GPSSMs has not been satisfactorily addressed yet:system identification. To address this challenge, we impose a structuredGaussian variational posterior distribution over the latent states, which isparameterised by a recognition model in the form of a bi-directional recurrentneural network. Inference with this structure allows us to recover a posteriorsmoothed over the entire sequence(s) of data. We provide a practical algorithmfor efficiently computing a lower bound on the marginal likelihood using thereparameterisation trick. This additionally allows arbitrary kernels to be usedwithin the GPSSM. We demonstrate that we can efficiently generate plausiblefuture trajectories of the system we seek to model with the GPSSM, requiringonly a small number of interactions with the true system.
-
Conference paperRafiq Y, Dickens L, Russo A, et al., 2017,
Learning to share: engineering adaptive decision-support for online social networks
, 32nd IEEE/ACM International Conference on Automated Software Engineering (ASE), Publisher: IEEE, Pages: 280-285, ISSN: 1527-1366Some online social networks (OSNs) allow users to define friendship-groups as reusable shortcuts for sharing information with multiple contacts. Posting exclusively to a friendship-group gives some privacy control, while supporting communication with (and within) this group. However, recipients of such posts may want to reuse content for their own social advantage, and can bypass existing controls by copy-pasting into a new post; this cross-posting poses privacy risks. This paper presents a learning to share approach that enables the incorporation of more nuanced privacy controls into OSNs. Specifically, we propose a reusable, adaptive software architecture that uses rigorous runtime analysis to help OSN users to make informed decisions about suitable audiences for their posts. This is achieved by supporting dynamic formation of recipient-groups that benefit social interactions while reducing privacy risks. We exemplify the use of our approach in the context of Facebook.
-
Conference paperKamthe S, Deisenroth MP, 2017,
Data-Efficient Reinforcement Learning with Probabilistic Model Predictive Control
, International Conference on Artificial Intelligence and StatisticsTrial-and-error based reinforcement learning (RL) has seen rapid advancementsin recent times, especially with the advent of deep neural networks. However,the majority of autonomous RL algorithms either rely on engineered features ora large number of interactions with the environment. Such a large number ofinteractions may be impractical in many real-world applications. For example,robots are subject to wear and tear and, hence, millions of interactions maychange or damage the system. Moreover, practical systems have limitations inthe form of the maximum torque that can be safely applied. To reduce the numberof system interactions while naturally handling constraints, we propose amodel-based RL framework based on Model Predictive Control (MPC). Inparticular, we propose to learn a probabilistic transition model using GaussianProcesses (GPs) to incorporate model uncertainties into long-term predictions,thereby, reducing the impact of model errors. We then use MPC to find a controlsequence that minimises the expected long-term cost. We provide theoreticalguarantees for the first-order optimality in the GP-based transition modelswith deterministic approximate inference for long-term planning. The proposedframework demonstrates superior data efficiency and learning rates compared tothe current state of the art.
-
Journal articleArulkumaran K, Deisenroth MP, Brundage M, et al., 2017,
A brief survey of deep reinforcement learning
, IEEE Signal Processing Magazine, Vol: 34, Pages: 26-38, ISSN: 1053-5888Deep reinforcement learning (DRL) is poised to revolutionize the field of artificial intelligence (AI) and represents a step toward building autonomous systems with a higherlevel understanding of the visual world. Currently, deep learning is enabling reinforcement learning (RL) to scale to problems that were previously intractable, such as learning to play video games directly from pixels. DRL algorithms are also applied to robotics, allowing control policies for robots to be learned directly from camera inputs in the real world. In this survey, we begin with an introduction to the general field of RL, then progress to the main streams of value-based and policy-based methods. Our survey will cover central algorithms in deep RL, including the deep Q-network (DQN), trust region policy optimization (TRPO), and asynchronous advantage actor critic. In parallel, we highlight the unique advantages of deep neural networks, focusing on visual understanding via RL. To conclude, we describe several current areas of research within the field.
-
Conference paperChabierski P, Russo A, Law M, et al., 2017,
Machine comprehension of text using combinatory categorial grammar and answer set programs
, COMMONSENSE 2017, Publisher: CEUR Workshop Proceedings, ISSN: 1613-0073We present an automated method for generating Answer Set Programs from narratives written in English and demonstrate how such a representation can be used to answer questions about text. The proposed approach relies on a transparent interface between the syntax and semantics of natural language provided by Combinatory Categorial Grammars to translate text into Answer Set Programs, hence creating a knowledge base that, together with background knowledge, can be queried.
-
Conference paperBaroni P, Comini G, Rago A, et al., 2017,
Abstract Games of Argumentation Strategy and Game-Theoretical Argument Strength
, PRIMA, Publisher: Springer, Pages: 403-419, ISSN: 0302-9743We define a generic notion of abstract games of argumentation strategy for (attack-only and bipolar) argumentation frameworks, which are zero-sum games whereby two players put forward sets of arguments and get a reward for their combined choices. The value of these games, in the classical game-theoretic sense, can be used to define measures of (quantitative) game-theoretic strength of arguments, which are different depending on whether either or both players have an “agenda” (i.e. an argument they want to be accepted). We show that this general scheme captures as a special instance a previous proposal in the literature (single agenda, attack-only frameworks), and seamlessly supports the definition of a spectrum of novel measures of game-theoretic strength where both players have an agenda and/or bipolar frameworks are considered. We then discuss the applicability of these instances of game-theoretic strength in different contexts and analyse their basic properties.
-
Conference paperRago A, Toni F, 2017,
Quantitative Argumentation Debates with Votes for Opinion Polling
, PRIMA, Publisher: Springer, Pages: 369-385, ISSN: 0302-9743Opinion polls are used in a variety of settings to assess the opinions of a population, but they mostly conceal the reasoning behind these opinions. Argumentation, as understood in AI, can be used to evaluate opinions in dialectical exchanges, transparently articulating the reasoning behind the opinions. We give a method integrating argumentation within opinion polling to empower voters to add new statements that render their opinions in the polls individually rational while at the same time justifying them. We then show how these poll results can be amalgamated to give a collectively rational set of voters in an argumentation framework. Our method relies upon Quantitative Argumentation Debate for Voting (QuAD-V) frameworks, which extend QuAD frameworks (a form of bipolar argumentation frameworks in which arguments have an intrinsic strength) with votes expressing individuals’ opinions on arguments.
-
Journal articleZheng JX, Pawar S, Goodman DFM, 2017,
Graph Drawing by Stochastic Gradient Descent
A popular method of force-directed graph drawing is multidimensional scalingusing graph-theoretic distances as input. We present an algorithm to minimizeits energy function, known as stress, by using stochastic gradient descent(SGD) to move a single pair of vertices at a time. Our results show that SGDcan reach lower stress levels faster and more consistently than majorization,without needing help from a good initialization. We then show how the uniqueproperties of SGD make it easier to produce constrained layouts than previousapproaches. We also show how SGD can be directly applied within the sparsestress approximation of Ortmann et al. [1], making the algorithm scalable up tolarge graphs.
-
Conference paperBao Z, Cyras K, Toni F, 2017,
ABAplus: Attack Reversal in Abstract and Structured Argumentation with Preferences
, PRIMA 2017: The 20th International Conference on Principles and Practice of Multi-Agent Systems, Publisher: Springer Verlag, ISSN: 0302-9743We present ABAplus, a system that implements reasoningwith the argumentation formalism ABA+. ABA+ is a structured argumentationformalism that extends Assumption-Based Argumentation(ABA) with preferences and accounts for preferences via attack reversal.ABA+ also admits as instance Preference-based Argumentation whichaccounts for preferences by reversing attacks in abstract argumentation(AA). ABAplus readily implements attack reversal in both AA and ABAstylestructured argumentation. ABAplus affords computation, visualisationand comparison of extensions under five argumentation semantics.It is available both as a stand-alone system and as a web application.
-
Conference paperOlofsson S, Mehrian M, Geris L, et al., 2017,
Bayesian multi-objective optimisation of neotissue growth in a perfusion bioreactor set-up
, European Symposium on Computer Aided Process Engineering (ESCAPE 27), Publisher: ElsevierWe consider optimising bone neotissue growth in a 3D scaffold during dynamic perfusionbioreactor culture. The goal is to choose design variables by optimising two conflictingobjectives: (i) maximising neotissue growth and (ii) minimising operating cost. Our con-tribution is a novel extension of Bayesian multi-objective optimisation to the case of oneblack-box (neotissue growth) and one analytical (operating cost) objective function, thathelps determine, within a reasonable amount of time, what design variables best managethe trade-off between neotissue growth and operating cost. Our method is tested againstand outperforms the most common approach in literature, genetic algorithms, and showsits important real-world applicability to problems that combine black-box models witheasy-to-quantify objectives like cost.
-
Journal articleHerrero P, Bondia J, Oliver N, et al., 2017,
A coordinated control strategy for insulin and glucagon delivery in type 1 diabetes
, Computer Methods in Biomechanics and Biomedical Engineering, Vol: 20, Pages: 1474-1482, ISSN: 1025-5842Type 1 diabetes is an autoimmune condition characterised by a pancreatic insulin secretion deficit, resulting in high blood glucose concentrations, which can lead to micro- and macrovascular complications. Type 1 diabetes also leads to impaired glucagon production by the pancreatic α-cells, which acts as a counter-regulatory hormone to insulin. A closed-loop system for automatic insulin and glucagon delivery, also referred to as an artificial pancreas, has the potential to reduce the self-management burden of type 1 diabetes and reduce the risk of hypo- and hyperglycemia. To date, bihormonal closed-loop systems for glucagon and insulin delivery have been based on two independent controllers. However, in physiology, the secretion of insulin and glucagon in the body is closely interconnected by paracrine and endocrine associations. In this work, we present a novel biologically-inspired glucose control strategy that accounts for such coordination. An in silico study using an FDA-accepted type 1 simulator was performed to evaluate the proposed coordinated control strategy compared to its non-coordinated counterpart, as well as an insulin-only version of the controller. The proposed coordinated strategy achieves a reduction of hyperglycemia without increasing hypoglycemia, when compared to its non-coordinated counterpart.
-
Conference paperCocarascu O, Toni F, 2017,
Identifying attack and support argumentative relations using deep learning
, 2017 Conference on Empirical Methods in Natural Language Processing, Publisher: Association for Computational Linguistics, Pages: 1374-1379We propose a deep learning architecture tocapture argumentative relations ofattackandsupportfrom one piece of text to an-other, of the kind that naturally occur ina debate. The architecture uses two (uni-directional or bidirectional) Long Short-Term Memory networks and (trained ornon-trained) word embeddings, and al-lows to considerably improve upon exist-ing techniques that use syntactic featuresand supervised classifiers for the sameform of (relation-based) argument mining.
-
Journal articleBiffi C, Simoes Monteiro de Marvao A, Attard M, et al., 2017,
Three-dimensional Cardiovascular Imaging-Genetics: A Mass Univariate Framework
, Bioinformatics, ISSN: 1367-4803Motivation: Left ventricular (LV) hypertrophy is a strong predictor of cardiovascular outcomes, but its genetic regulation remains largely unexplained. Conventional phenotyping relies on manual calculation of LV mass and wall thickness, but advanced cardiac image analysis presents an opportunity for highthroughput mapping of genotype-phenotype associations in three dimensions (3D).Results: High-resolution cardiac magnetic resonance images were automatically segmented in 1,124 healthy volunteers to create a 3D shape model of the heart. Mass univariate regression was used to plot a 3D effect-size map for the association between wall thickness and a set of predictors at each vertex in the mesh. The vertices where a significant effect exists were determined by applying threshold-free cluster enhancement to boost areas of signal with spatial contiguity. Experiments on simulated phenotypic signals and SNP replication show that this approach offers a substantial gain in statistical power for cardiac genotype-phenotype associations while providing good control of the false discovery rate. This framework models the effects of genetic variation throughout the heart and can be automatically applied to large population cohorts.Availability: The proposed approach has been coded in an R package freely available at https://doi.org/10.5281/zenodo.834610 together with the clinical data used in this work.
-
Conference paperChamberlain BP, Cardoso A, Liu CHB, et al., 2017,
Customer lifetime value pediction using embeddings
, International Conference on Knowledge Discovery and Data Mining, Publisher: ACM, Pages: 1753-1762We describe the Customer LifeTime Value (CLTV) prediction system deployed at ASOS.com, a global online fashion retailer. CLTV prediction is an important problem in e-commerce where an accurate estimate of future value allows retailers to effectively allocate marketing spend, identify and nurture high value customers and mitigate exposure to losses. The system at ASOS provides daily estimates of the future value of every customer and is one of the cornerstones of the personalised shopping experience. The state of the art in this domain uses large numbers of handcrafted features and ensemble regressors to forecast value, predict churn and evaluate customer loyalty. Recently, domains including language, vision and speech have shown dramatic advances by replacing handcrafted features with features that are learned automatically from data. We detail the system deployed at ASOS and show that learning feature representations is a promising extension to the state of the art in CLTV modelling. We propose a novel way to generate embeddings of customers, which addresses the issue of the ever changing product catalogue and obtain a significant improvement over an exhaustive set of handcrafted features.
-
Journal articleHerrero P, Bondia J, Adewuyi O, et al., 2017,
Enhancing automatic closed-loop glucose control in type 1 diabetes with an adaptive meal bolus calculator - in silico evaluation under intra- day variability
, Computer Methods and Programs in Biomedicine, Vol: 146, Pages: 125-131, ISSN: 0169-2607Background and ObjectiveCurrent prototypes of closed-loop systems for glucose control in type 1 diabetes mellitus, also referred to as artificial pancreas systems, require a pre-meal insulin bolus to compensate for delays in subcutaneous insulin absorption in order to avoid initial post-prandial hyperglycemia. Computing such a meal bolus is a challenging task due to the high intra-subject variability of insulin requirements. Most closed-loop systems compute this pre-meal insulin dose by a standard bolus calculation, as is commonly found in insulin pumps. However, the performance of these calculators is limited due to a lack of adaptiveness in front of dynamic changes in insulin requirements. Despite some initial attempts to include adaptation within these calculators, challenges remain.MethodsIn this paper we present a new technique to automatically adapt the meal-priming bolus within an artificial pancreas. The technique consists of using a novel adaptive bolus calculator based on Case-Based Reasoning and Run-To-Run control, within a closed-loop controller. Coordination between the adaptive bolus calculator and the controller was required to achieve the desired performance. For testing purposes, the clinically validated Imperial College Artificial Pancreas controller was employed. The proposed system was evaluated against itself but without bolus adaptation. The UVa-Padova T1DM v3.2 system was used to carry out a three-month in silico study on 11 adult and 11 adolescent virtual subjects taking into account inter-and intra-subject variability of insulin requirements and uncertainty on carbohydrate intake.ResultsOverall, the closed-loop controller enhanced by an adaptive bolus calculator improves glycemic control when compared to its non-adaptive counterpart. In particular, the following statistically significant improvements were found (non-adaptive vs. adaptive). Adults: mean glucose 142.2 ± 9.4 vs. 131.8 ± 4.2 mg/dl; perce
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.