Staff Profiles

Supervision

Solomon, B. 2015 – present, A Framework for Modelling Social Media Collaborative Work
Sturgess, P. 2014 – 2016, Holistic Street Scene Understanding as a Higher Order Labelling Problem with Associative Hierarchical Random Fields, PhD awarded 2016.
Singh, G. 2015 – present, Continuous human action detection and future action prediction.
Saha, S. 2014 – present, Online Multiple Human Action Detection from Videos Submitted Jan 2018
Lee, M. 2014 – present, 3D action recognition in a real-time online environment, transferred to University of Oxford
Lidegaard, M. 2013-2014, Computer Vision and Natural Language to Help the Partially Sighted, transferred to University of Oxford.
Barker, S. 2011-present (part-time), Design, build and experimentation with a humaniod head and neck robot. PhD awarded 2017.
Sheasby, G. 2009-2013, Beyond Controllers: Object Detection and Human Pose Estimation as Game Input Mechanisms, PhD awarded 2013
Rihan, J. 2006–2011, Computer Vision for Computer Games, PhD awarded 2011

Goh, W.J. 2003–2008, Some Computational properties of Chaotic Neural Networks,PhD awarded 2008.

Howarat, M., 2001–2007, Adaptive Chaotic Neural Networks, PhD awarded 2007.

Tator, S.,1998 - 2001, Supporting Adaptive Continuous Media over Packet Switched Networks, PhD awarded 2002.

olde Scheper, T. 1997–2000, Chaos and information in Dynamic Neural Networks, PhD awarded 2002.

Howes, P. 1995 - 1999, Automated Explanation Facilities for Neural Networks, PhD awarded 1999.

McQuoid, M.R.J.,1992-1995, A Neural Architecture for Emergent Serial Behaviour, PhD awarded 1996.

Howes, P. 1990-1993, The Application of Expert Systems to Forecasting with Time Series Data, MPhil awarded 1993.

My research interests include Ethical Artificial Intelligence, autonomous moral machiens, biologically inspired machine learning, embodied conversational agents and social robotics. In my most recent research I have been exploring how it might be possible to equip robots with a degree of moral competence. This would enable robots to recognise the moral consequences of their actions and be able to predict what kinds of behaviours are acceptable in a given situation. I am also working on other aspects of ethical AI, including developing systems that can explain the decision of trained machine learning models and automatic personality typing. I am founder of the Ethical AI institute at Oxford Brookes University (formal launch summer 2019). My work on robotics has attracted some media attention, including 16 appearances on regional, national and international television channels.

Research group membership

Cognitive Robotics Laboratory (Head)

Artificial Intelligence and Vision.

Research grants and awards

"Improving legal services productivity and accessibility with Artificial Intelligence" - Innovate UK - (PI:N. Crook , Partners - Moorcroft Ltd, awarded 2019)

WEKIT – "Wearable Experience for Knowledge Intensive Training" - EU H2020 – (PI: F. Wild – 13 Partners - transferred to OBU 2015).

TCBL – "Textile & Clothing Business Labs" - EU H2020: (PI: F. Wild – transferred to OBU 2015).

"Application of computer vision robotic welds" - Innovate UK (TSB) KTP (Lead Academic: N.T. Crook, Academic Supervisor: F. Cuzzolin, Partner: Meta Vision Systems. Awarded March 2015.

Further information

robotics.ecm.brookes.ac.uk/

Journal articles

• Aldea A, Crook N, Duce D, Marshall P, Martin C, Sutton D, 'Reflections on the evolution of the teaching of programming to undergraduates at Oxford Brookes University'
  Brookes eJournal of Learning and Teaching 7 (1) (2015)
  ISSN: 1744-7747
  Abstract

  Undergraduate computing courses inevitably include a high degree of regeneration in order to keep abreast of this rapidly changing field. Introductory programming modules in particular need to adapt to changing trends and languages. Until recently, the focus of debate within the Oxford Brookes University curriculum has therefore been on the course content, but since 2012 there has been a major change in the method of delivery through the introduction of a new apprenticeship model. This paper seeks to reflect on this, and other recent changes which have led to improved student engagement and results. The data is limited however, and so the results presented here are not conclusive.

  Website 
• Cheng M, Zheng S, Lin WY, Vineet V, Sturgess P, Crook N, Mitra NJ, Torr P, 'Imagespirit:verbal guided image parsing'
  ACM Transactions on Graphics 34 (1) (2014)
  ISSN: 0730-0301
  Abstract

  Humans describe images in terms of nouns and adjectives while algorithms

  operate on images represented as sets of pixels. Bridging this gap between

  how humans would like to access images versus their typical representation

  is the goal of image parsing, which involves assigning object and attribute

  labels to pixel. In this paper we propose treating nouns as object labels and

  adjectives as visual attribute labels. This allows us to formulate the image

  parsing problem as one of jointly estimating per-pixel object and attribute

  labels from a set of training images. We propose an efficient (interactive

  time) solution. Using the extracted labels as handles, our system empowers

  a user to verbally refine the results. This enables hands-free parsing of an

  image into pixel-wise object/attribute labels that correspond to human semantics.

  Verbally selecting objects of interests enables a novel and natural

  interaction modality that can possibly be used to interact with new generation

  devices (e.g. smart phones, Google Glass, living room devices). We

  demonstrate our system on a large number of real-world images with varying

  complexity. To help understand the tradeoffs compared to traditional

  mouse based interactions, results are reported for both a large scale quantitative

  evaluation and a user study.

  Website 
• Crook N, Field D, Smith C, Harding S, Pulman S, Cavazza M, Moore R, Boye J, 'Generating context-sensitive ECA responses to user barge-in interruptions'
  Journal on Multimodal User Interfaces 6 (40940) (2012) pp.13-25
  ISSN: 1783-7677
  Abstract

  We present an Embodied Conversational Agent(ECA) that incorporates a context-sensitive mechanism for handling user barge-in. The affective ECA engages the user in social conversation, and is fully implemented. We will use actual examples of system behaviour to illustrate. The ECA is designed to recognise and be empathetic to the emotional state of the user. It is able to detect, react quickly to, and then follow up with considered responses to different kinds of user interruptions. The design of the rules which enable the ECA to respond intelligently to different types of interruptions was informed by manually analysed real data from human-human dialogue. The rules represent recoveries from interruptions as two-part structures: an address followed by a resumption. The system is robust enough to man- age long, multi-utterance turns by both user and system, which creates good opportunities for the user to interrupt while the ECA is speaking.

  Website 
• Smith C, Crook N, Charlton D, Boye J, Pulman S, Santos de la Camara R, Turunen M, Benyon D, 'Interaction Strategies for an Affective Conversational Agent'
  Presence: Teleoperators and Virtual Environments 20 (5) (2012)
  ISSN: 1054-7460
  Website 
• Crook N, Goh W, 'Nonlinear transient computation as a potential 'kernel trick' in cortical processing'
  BioSystems 94 (1) (2008) pp.55-59
  ISSN: 0303-2647 eISSN: 1872-8324
  Abstract Evidence has been found for the presence of chaotic dynamics at all levels of the mammalian brain. This has led to some searching questions about the potential role that nonlinear dynamics may have in neural information processing. We propose that chaos equips the brain with the equivalent of a kernel trick for solving hard nonlinear problems. The approach presented, which is described as nonlinear transient computation, uses the dynamics of a well known chaotic attractor. The paper provides experimental results to show that this approach can be used to solve some challenging pattern recognition tasks. The paper also offers evidence to suggest that the efficacy of nonlinear transient computation for nonlinear pattern classification is dependent only on the generic properties of chaotic attractors and is not sensitive to the particular dynamics of specific sub-regions of chaotic phase space. If, as this work suggests, nonlinear transient computation is independent of the particulars of any given chaotic attractor, then it could be offered as a possible explanation of how the chaotic dynamics that have been observed in brain structures contribute to neural information processing tasks.Website 
• Crook N, 'Nonlinear Transient Computation'
  Neurocomputing 70 (2007) pp.1167-1176
  Website 
• Crook N, Scheper TO, Pathirana V, 'Self-organised Dynamic Recognition States for Chaotic Neural Networks'
  Information Sciences 150 (1-2) (2003) pp.59-75
  ISSN: 0020-0255
  Abstract Chaos offers several advantages to the Engineer over other non-chaotic dynamics. One is that chaotic systems are often significantly easier to control than other linear or non-linear systems, requiring only small, appropriately timed perturbations to constrain them within specific unstable periodic orbits (UPOs). Another is that chaotic attractors contain an infinite number of these UPOs. If individual UPOs can be made to represent specific internal states of a system, then a chaotic attractor can be turned into an infinite state machine. In this paper we investigate this possibility with respect to chaotic neural networks. We present a method by which a network can self-select UPOs in response to specific input values. These UPOs correspond to network recognition states for these input values.Website 
• Crook NT, olde Scheper T, Pathirana V, 'Self organised dynamic recognition states for chaotic neural networks'
  Information Sciences 150 (1-2) (2003) pp.59-75
  ISSN: 0020-0255
  Abstract Chaos offers several advantages to the Engineer over other non-chaotic dynamics. One is that chaotic systems are often significantly easier to control than other linear or non-linear systems, requiring only small, appropriately timed perturbations to constrain them within specific unstable periodic orbits (UPOs). Another is that chaotic attractors contain an infinite number of these UPOs. If individual UPOs can be made to represent specific internal states of a system, then a chaotic attractor can be turned into an infinite state machine. In this paper we investigate this possibility with respect to chaotic neural networks. We present a method by which a network can self-select UPOs in response to specific input values. These UPOs correspond to network recognition states for these input values.Website 
• Crook NT, olde Scheper T, 'Adaptation based on memory dynamics in a chaotic neural network'
  Cybernetics and Systems 33 (4) (2002) pp.341-378
  ISSN: 0196-9722
  Abstract The complex dynamics that emerge from systems governed by deterministic chaos offer significant advantages to the neuromorphic engineer. Included in these is the potential for a very large memory store and the ease with which chaotic systems can be controlled. By definition, a chaotic system is a periodic. However, during the course of its trajectory through state space, the chaotic system will come infinitely close to points that it has previously visited. These almost repeating trajectories are referred to as Unstable Periodic Orbits (UPOs). Normally, under the influence of chaos, the trajectory would move away exponentially fast from its previous path, thereby describing a new path on the surface of the attractor. It is possible to apply a simple delayed feedback control mechanism to a chaotic system that will constrain it within one of its UPOs. This article presents a neural implementation of this delayed feedback mechanism. The network presented here is able to stabilize different UPOs in response to different input signals, with each UPO corresponding to a dynamic recognition state for that input. We also present two learning rules for this network, which enables it to adapt to novel inputs in a self-organized manner.Website 
• Howes P, Crook N, 'Using Input Parameter Influences to Support the Decisions of Feedforward Neural Networks'
  Neurocomputing 24 (1999) pp.191-206
  

Book chapters

• Beton L, Hughes P, Barker S, Pilling M, Fuente L, Crook NT, 'Leader-Follower Strategies for Robot-Human Collaboration' in Aldinhas Ferreira MA, Silva Sequeira J, Osman Tokhi M, Kadar EE, Singh Virk G (ed.), A World with Robots, Springer (2017)
  ISBN: 9783319466651 eISBN: 9783319466675
  Abstract This paper considers the impact that robot collaboration strategies have on their human collaborators. In particular, we are interested in how robot leader/follower strategies affect perceived safety and perceived intelligence, which, we argue, are essential for establishing trust and enabling true collaboration between human and robot. We propose an experiment which will enable us to evaluate the impact of leader/follower collaboration strategies on perceived safety and intelligence.Website 
• Fuente LA, Lones MA, Crook NT, Olde Scheper TV, 'Harmonic versus chaos controlled oscillators in hexapedal locomotion' in Information Processing in Cells and Tissues, (2015)
  Abstract

  The behavioural diversity of chaotic oscillator can be controlled into periodic dynamics and used to model locomotion using central pattern generators. This paper shows how controlled chaotic oscillators may improve the adaptation of the robot locomotion behaviour to terrain uncertainties when compared to nonlinear harmonic oscillators. This is quantitatively assesses by the stability, changes of direction and steadiness of the robotic movements. Our results show that the controlled Wu oscillator promotes the emergence of adaptive locomotion when deterministic sensory feedback is used. They also suggest that the chaotic nature of chaos controlled oscillators increases the expressiveness of  pattern generators to explore new locomotion gaits.

  Website 
• Crook N, Dobbyn C, olde Scheper T, 'Chaos as a Desirable Stable State of Artificial Neural Networks' in Advances in Soft Computing: Soft Computing Techniques and Applications, Physica-Verlag (2000)
  

Conference papers

• Rolf M, Crook N, Steil J, 'From Social Interaction to Ethical AI: A Developmental Roadmap'
  (2018)
  eISSN: 2161-9484
  Abstract
  AI and robot ethics have recently gained a lot of attention because adaptive machines are increasingly involved in ethically sensitive scenarios and cause incidents of public outcry. Much of the debate has been focused on achieving highest moral standards in handling ethical dilemmas on which not even humans can agree, which indicates that the wrong questions are being asked. We suggest to address this ethics debate strictly through the lens of what behavior seems socially acceptable, rather than idealistically ethical. Learning such behavior puts the debate into the very heart of developmental robotics. This paper poses a roadmap of computational and experimental questions to address the development of socially acceptable machines. We emphasize the need for social reward mechanisms and learning architectures that integrate these while reaching beyond limitations of plain reinforcement learning agents. We suggest to use the metaphor of “needs” to bridge rewards and higher level abstractions such as goals for both communication and action generation in a social context. We then suggest a series of experimental questions and possible platforms and paradigms to guide future research in the area.
  Website 
• Barker S, Izadi H, Crook NT, Hayatleh K, Rolf M, Hughes P, Fellows N, 'Natural head movement for HRI with a muscular-skeletal head and neck robot'
  (2017) pp.587-592
  eISSN: 1944-9437 ISBN: 9781538635186
  Abstract This paper presents a study of the movements of a humanoid head-and-neck robot called Eddie. Eddie has a musculo-skeletal structure similar to that found in human necks enabling it to perform head movements that are comparable with human head movements. This study compares the movements of Eddie with those of a more conventional robotic neck structure and with those of a human head. Results show that Eddie’s movements are perceived as significantly more natural and by trend more lifelike than the conventional head’s. No differences were found with respect to the impression of humanlikeness, consciousness, and elegance.
   
  Website 
• Rolf M, Crook NT, 'What if: robots create novel goals? Ethics based on social value systems'
  1668 (2016) pp.20-25
  Abstract Future personal robots might possess the capability to autonomously generate novel goals that exceed their initial programming as well as their past experience. We discuss the ethical challenges involved in such a scenario, ranging from the construction of ethics into such machines to the standard of ethics we could actually demand from such machines. We argue that we might have to accept those machines committing human-like ethical failures if they should ever reach human-level autonomy and intentionality. We base our discussion on recent ideas that novel goals could be originated from agents’ value system that express a subjective goodness of world or internal states. Novel goals could then be generated by extrapolating what future states would be good to achieve. Ethics could be built into such systems not just by simple utilitarian measures but also by constructing a value for the expected social acceptance of a the agent’s conduct.
   
  Website 
• Fuente LA, Ierardi H, Pilling M, Crook NT, 'Influence of Upper Body Pose Mirroring in Human-Robot Interaction'
  9388 (2015) pp.214-223
  eISBN: 978-3-319-25554-5
  Abstract This paper explores the effect of upper body pose mirroring in human-robot interaction. A group of participants is used to evaluate how imitation by a robot affects people’s perception of their conversation with it. A set of twelve questions about the participants’ university experience serves as a backbone for the dialogue structure. In our experimental evaluation, the robot reacts in one of three ways to the human upper body pose: ignoring it, displaying its own upper body pose, and mirroring it. The manner in which the robot behaviour influences human appraisal is analysed using the standard Godspeed questionnaire. Our results show that robot body mirroring/non-mirroring influences the perceived humanness of the robot. The results also indicate that body pose mirroring is an important factor in facilitating rapport and empathy in human social interactions with robots.Website 
• Barker S, Fuente L, Hayatleh K, Fellows N, Steil JJ, Crook N, 'Design of a Biologically Inspired Humanoid Neck'
  (15820238) (2015) pp.25-30
  ISBN: 978-1-4673-9674-5 eISBN: 978-1-4673-9675-2
  Abstract This paper presents the design of a novel anthropomorphic robotic neck. It mimics the range of movements found in the human neck, actuated by pneumatic artificial muscles. The proposed humanoid neck simulates the anatomical functionality and structure of a human neck. Specifications are made according to biological, anatomical and behavioural data. The preliminary results show that the proposed humanoid neck is able to deliver the range of movements and head velocities comparable to those observed in human necks. These results also demonstrate that biologically inspired musculoskeletal robotic systems represent a reliable and robust platform to investigate motion development.Website 
• Cheng MM, Warrell J, Lin WY, Zheng S, Vineet V, Crook N, 'Efficient salient region detection with soft image abstraction'
  (2013) pp.1529-1536
  ISBN: 9781479928408
  Abstract

  Detecting visually salient regions in images is one of the fundamental problems in computer vision. We propose a novel method to decompose an image into large scale perceptually homogeneous elements for efficient salient region detection, using a soft image abstraction representation. By considering both appearance similarity and spatial distribution of image pixels, the proposed representation abstracts out unnecessary image details, allowing the assignment of comparable saliency values across similar regions, and producing perceptually accurate salient region detection. We evaluate our salient region detection approach on the largest publicly available dataset with pixel accurate annotations. The experimental results show that the proposed method outperforms 18 alternate methods, reducing the mean absolute error by 25.2% compared to the previous best result, while being computationally more efficient.
  

  Website 
• Sheasby G, Warrell J, Zhang Y, Crook N, Torr P, 'Simultaneous human segmentation, depth and pose estimation via dual decomposition'
  (2012) pp.1-12
  Abstract The tasks of stereo matching, segmentation, and human pose estimation have been popular in computer vision in recent years, but attempts to combine the three tasks have so far resulted in compromises: either using infra-red cameras, or a greatly simplified body model. We propose a framework for estimating a detailed human skeleton in 3D from a stereo pair of images. Within this framework, we define an energy function that incorporates the relationship between the segmentation results, the pose estimation results, and the disparity space image. Specifically, we codify the assertions that foreground pixels should relate to some body part, should correspond to a continuous surface in the disparityspace image, and should be closer to the camera than the surrounding background pixels. Our energy function is NP-hard, however we show how to efficiently optimize a relaxation of it using dual decomposition. We show that applying this approach leads to improved results in all three tasks, and also introduce an extensive and challenging new dataset, which we use as a benchmark for evaluating 3D human pose estimation.
• Zhang Z, Sturgess P, Sengupta S, Crooke N, Torr P, 'Efficient discriminative learning of parametric nearest neighbor classifiers'
  (2012) pp.2232-2239
  Website 
• Crook N, Goh WJ, Hawarat M, 'Pattern Recall in Networks of Chaotic Neurons'
  Biosystems 87 (2007) pp.267-274
  ISSN: 0303-2647 eISSN: 1872-8324
  Website 
• Crook N, Scheper TO, 'An Adaptive Chaotic Neural Network'
  Proceeding of the 2002 International Joint Conference on Neural Networks, Vols 1-3 (2002) pp.2580-2585
  
• Crook NT, olde Scheper T, 'Learning in a Chaotic Neural Network'
  (2002) pp.401-406
  ISBN: 2-930307-02-1
  Website 
• olde Scheper T, Crook NT, 'Delay Control Aspects of the Chaotic Rossler System'
  (2002)
  ISBN: 3-906454-32-0
  Website 
• Crook NT, olde Scheper T, 'Adaptive Memory Dynamics in a Chaotic Neural Network'
  33 (4) (2002) pp.131-
  ISBN: 3-906454-32-0
  Website 
• Crook NT, olde Scheper T, 'A novel chaotic neural network architecture'
  (2001) pp.295-300
  ISBN: 2930307013
  
• Crook N, Scheper TO, 'Dynamic Recognition States for Chaotic Neural Networks'
  Developments in Soft Computing (2001) pp.62-72
  Website 
• Crook N, Dobbyn C, Scheper TO, 'Chaos As a Desirable Stable State of Artificial Neural Networks'
  Soft Computing Techniques and Applications (2000) pp.52-60
  

Other publications

• Crook N, Scheper TO, 'Special Edition of Biosystems: Information Processing in Cells and Tissues'
  Biosystems 94 (2008) pp.1-1ISSN: 0303-2647 eISSN: 1872-8324
  Website 

Areas of expertise

Ethical AI (Autonomous Moral Machines, Explainable AI), Social Robotics, Machine Learning.

Membership of professional bodies

Member of IEEE

Principal Fellow of the Higher Education Academy

Press, publicity and reviews

BBC Radio Oxford 26^th February 2018 – Howard Bentham – 1^st public road testing of CAV to be in Oxfordshire
BBC Radio Oxford 9^th February 2018 – David Prever show – Comments on Autonomous moral vehicles
Premier Radio – 23^rd January 2018 - Video interview for Unbelievable program – to be made public in May 2018
BBC South Today – 24^th January 2017 – with Stephen Metcalfe MP (chair Commons Science & Tech com) visit
BBC Word News – 14^th September 2015 – Cognitive robotics research
BBC World News, Global programme – 6^th November 2014 – Realistic humanoid robots and computer graphics
BBC Radio Oxford – 24^th June 2014 – Robot Sumo Competition
BBC Radio Oxford – 27^th June 2014 – Robot Sumo Competition
BBC South Today – 27^th June 2014 – Robot Sumo Competition
CNBC Squawk Box – 1^st May 2014 – RoboThespian the Humanoid Robot
BBC South Today – 14^th March 2014 – Artie at the Ashmolean Museum, Oxford
ITV Meridian News – 13^th March 2014 - Artie the talking robot delights students at Oxford Brookes University
BBC Radio Oxford – 12^th March 2014 – Artie the humanoid robot
BBC Breakfast – 10^th March 2014 – RoboThespian joins BBC Breakfast
CBBC Newsround – 10th March 2014 – Newsround chats with robot expert