The Cognition Institute is pleased to announce that many of the external seminars hosted by research groups within the institute will now be taking place in a single venue – the workshop space in Link 3. The space will host weekly and monthly seminars by Transtechnology, School of Psychology, CRNS and CogNovo amongst others. The series continued for another season.
This Transtechnology Research seminar series explores how an object can be understood as a nexus of discourses, and how the arranging of those discourses in relation to a particular theoretical framework can offer reciprocal insights into the object and the framework. It also offers some traction in the ongoing discussion of cognition and media (practices). Each seminar will have as its primary discursive focus an object and through this, aspects of theoretical and practical research will find engagement with an extended realm of enquiry. In this series each move will be driven by the discussion of the object either as a symptom, mnemonic or dream of an affective and affectionate media interaction; oblique strategies to engage with the avant-garde and recover art as an aesthetic experience that allows the artist and the artwork to affectively connect with whoever encounters it.
The series follows on from the series this year and explores the affective relationship of and with objects, media and technologies. We have called the series Objects of Affect and Affection in order to take forward our argument and examine the effects and consequences of the tendency of digitization to flatten difference.
23/09/2015; 1pm; Link3 workspace area
Transtechnology Research seminar hosted by: Martha Blassnigg, Hannah Drayson, and Michael Punt
James Marshall (University of Sheffield)
Effective decision-making is crucial for organisms at all levels of biological complexity. I will present a model of collective decision-making based on empirical observations of a novel cross-inhibitory behaviour in house-hunting honeybee swarms. The pattern of interactions observed in collectively-deciding honeybees gives rise to a number of important value-sensitive decision-making characteristics. The model is able to achieve stable deadlock for poor but equal alternatives, but spontaneously choose between good alternatives. This enables sophisticated 'wait and see' decision-making. The model's sensitivity to value is similar to Weber's law of just-noticable-difference from psychology. When differences are large enough to be noticeable, the model exhibits speed-accuracy trade-offs in decision-making similar to classic models from psychology. Given the simplicity of the model, the importance of value-sensitivity, and the similar patterns of interaction seen in other decision-making systems, I will speculate about whether other natural systems may implement the same algorithms, and discuss the application of these to the design of decentralised algorithms for very large robot swarms.
2/10/2015; 2pm; Link3 workspace area
CRNS Seminar hosted by: Roman Borisyuk, Robert Merrison and Martin F. Stoelen
Luís Seabra Lopes (Universidade de Aveiro, Portugal)
An overview of research on open-ended learning in robotics is presented. Key characteristics of intelligent service robots as well as some of the issues in the development of such robots are identified. The presentation focuses on two important phases in experience-based learning, namely experience extraction and experience conceptualization. These two learning steps are addressed in two different domains, namely object category learning and activity schema learning. The human user, playing the role of instructor, helps to speed up and focus the learning process. Aspects of evaluation of open-ended learning are also addressed.
9/10/2015; 2pm; Link3 workspace area
CRNS Seminar hosted by: Guido Bugmann and Martin F. Stoelen
Stephen Hall (Plymouth University)
HBrain rhythms or ‘Oscillations’ are neuronal network phenomena, first recorded almost a century ago. In the time since these first recordings, brain rhythms have been studied across a wide range of species, under many different experimental conditions. Here, I will introduce the topic of brain rhythms, through a discussion of the various cognitive and behavioural functions in which they have been implicated. I will describe some of the basic physiological principles of oscillations and how this relates to our ability to measure them. I will discuss some of the differences between evoked and induced oscillations. Finally, I will explore some of the theories surrounding the potential significance and importance of these phenomena (or epiphenomena?).
21/10/2015; 4pm; Link3 workspace area
Psychology Seminar
Tadahiro Taniguchi (Ritsumeikan University, Japan)
Humans can acquire language through physical interaction with their environment and semiotic interaction with other people. It is very important to understand computationally how humans can form a symbol system and obtain semiotic skills through their autonomous mental development. We are challenging to construct a robotic system and machine learning method which can obtain language through embodied multimodal interaction. In this talk, I introduce the basis of our research and state-of-art results. Specifically, I talk about multimodal categorization and double articulation analyzer which enable a robot to obtain words and their embodied meanings from raw sensory-motor information and acoustic speech signals totally in an unsupervised manner.
23/10/2015; 2pm; Link3 workspace area
CRNS Seminar hosted by: Angelo Cangelosi and Martin F. Stoelen
Fred Cummins (University College Dublin)
Joint speech is an umbrella term covering choral speech, synchronous speech, chant, and all forms of speech where many people say the same thing at the same. Under an orthodox linguistic analysis, there is nothing here to study, as the formal symbolic structures of joint speech do not appear to differ from those of language arising in other forms of practice. As a result, there is essentially no body of scientific inquiry into practices of joint speaking. Yet joint speaking practices are ubiquitous, ancient, and deeply integrated into rituals and domains to which we accord the highest significance.
I will discuss Joint Speech, as found in prayer, protest, classrooms, and sports stadia around the world. If we merely take the time to look there is much to be found in joint speech that is crying out for elaboration and investigation. I will attempt to sketch the terra incognita that opens up and present a few initial findings (phonetic, anthropological, neuroscientific) that suggest that Joint Speech is far from being a peripheral and exotic special case. It is, rather, a central example of language use that must inform our theories of what language and languaging are.
4/11/2015; 4pm; Link3 workspace area
Psychology Research Seminar hosted by: Andy Willis
Helmut Hauser (University of Bristol)
Morphological computation describes conceptually the observation that biological systems take advantage of their morphology an its dynamic properties to conduct computations that are needed for a successful interaction with the environment. This results in systems that are more energy efficient, adaptive and resilient – all properties we want to see in robotic systems as well.
With the help of examples we will introduce the concept and show how it can be formalised in a mathematical framework, which is also able to describe how computation can be literally outsourced to the physical body of the agent. As a result the remaining computations and the corresponding learning and controlling tasks are much simpler. A remarkable conclusion of our theories is that soft, compliant bodies are computationally more powerful than classical, rigid robotic structure pointing to a great potential of soft robotics. Finally, we will discuss the range of possible applications of morphological computation by presenting state-of-the-art results in simulations as well on real robotic platforms and we will discuss the corresponding research opportunities ahead of us.
6/11/2015; 2pm; Link3 workspace area
CRNS Seminar hosted by: Angelo Cangelosi and Martin F. Stoelen
Clare Press (Birkbeck, University of London)
Mechanisms which map between the visual appearance of an action and the motor codes required to perform it are crucial for a range of functions, including imitation and action control, and possibly also play a role in action perception and understanding. The first part of my talk will present some studies addressing the domain-specificity of underlying mechanisms. It will examine whether the mechanisms mapping motor codes to observed actions are separable from those mapping motor codes to associated inanimate events, as required for stamping on the brake pedal when we see a red light. It will also investigate whether action influences perception of predicted sensory consequences in a different manner from inanimate predictive events. The second part of my talk will present work addressing differences in action production and perception in autism, and asking which mechanisms may be functioning atypically.
11/11/2015; 4pm; Link3 workspace area
Psychology Seminar hosted by: Patric Bach
Jochen Braun (Otto-von-Guericke-University Magdeburg)
Visual perception has all the hallmarks of an ongoing, cooperative-competitive process: probabilistic outcome, self-organization, order-disorder transitions, multi-stability, and hysteresis. It is therefore tempting to speculate that the underlying collective neural activity performs an exploratory attractor dynamics (spontaneous transitions between distinct steady-states), perhaps at multiple spatial and temporal scales. Here I summarize our recent investigations of this dynamical hypothesis. In several instances, a careful empirical study of perceptual dynamics fully constrains an idealized model of the stochastic dynamics of collective neural activity.
I conclude that the dynamical hypothesis outlined above permits a particularly close and direct back-and-forth between perceptual experiment and computational theory and thus has the potential to dramatically accelerate our progress in understanding visual function.
Related publications: Stochastic accumulation by cortical columns may explain the scalar property of multistable perception., Multi-stable perception balances stability and sensitivity., Believable change: bistable reversals are goverened by physical plausibility.
13/11/2015; 2pm; Link3 workspace area
CRNS Seminar hosted by: Sue Denham and Martin F. Stoelen
Jane Hutchinson
The purpose of this seminar is to explore the extent to which a re-creation or re-presentation of an object or event is able to ‘stand-in’ for the original, whether it was fully realised at the time of its conception or remained in an imaginary form.
It will do this by considering the provisional (re)-construction of Hugo Munsterberg’s apparatus for testing the "mental constitution" of motormen in order to determine their suitability for the job. The experiment took place at the Harvard Psychological Laboratory during the spring of 1912. The experimental subjects were employees of The Boston Elevated Railway Company. In his initial report Munsterberg provided a detailed account of his design of the experimental process, and the apparatus. He claimed its success was in part at least due to its evoking for his subjects the reality of their working lives.
The seminar is presented in the context of a wider exploration of the value of reconstruction in a media-archaeological research process. Drawings, images and a maquette model that were created during the research process will be presented. These are informed by Munsterberg’s description (see text below). Participants will be invited to engage with these items in order to explore the extent of their adequacy as ‘stand-ins’ for the apparatus, even if its original form was an imaginary constituent of a thought experiment.
More information about this seminar on the TransTechnology Research website.
18/11/2015; 1pm; Link3 workspace area
TransTechnology Seminar "Objects of Affect and Affection" hosted by: Hannah Drayson and Michael Punt
Stephanie Dornschneider (The University of Buckingham)
During mass uprisings, why do certain people join the protests against their governments, while others stay at home? Focusing on structural or organizational factors that contribute to political mobilization, much of the existing literature fails to address this difference in behavior. In response, this presentation draws on the literature on beliefs and belief systems to explore the reasoning processes by which individuals (fail to) decide to join political protests. Focusing on the Arab Spring as a particular case, it examines 121 protestors and non-protestors from Egypt - a country where the Arab Spring protests led to the fall of the president - and Morocco - a country where the head of state did not resign as a result of the uprisings. Information about the reasoning processes of these individuals was gathered through field research (ethnographic interviews) and Facebook groups. To construct reasoning processes from these sources, the analysis applied qualitative methods developed by Strauss and Corbin, coding the people's direct speech into beliefs, belief connections (inferences), and decisions for actions. To analyze these data, which consist of trillions of combinations of beliefs and inferences, the analysis developed a computational model (in Python). The model systematically evaluates the protestors' and non-protestors' reasoning processes, contributing new insight into the sources of political protest.
18/11/2015; 4pm; Link3 workspace area
Psychology Research Seminar hosted by: Matt Roser
Jack Mellor (University of Bristol)
The Nobel prize this year was awarded for the demonstration of neurons that encode location, place cells and grid cells. This has led to the idea that spatial memory is encoded within the network of these cells. One prediction is that synaptic plasticity between place cells reorganises the network to provide spatial representations of environments. We provide data to support this prediction and show how and when synaptic plasticity is induced, and therefore spatial memory encoded.
20/11/2015; 2pm; Link3 workspace area
CRNS Seminar hosted by: Guido Bugmann and Martin F. Stoelen
One of the biggest psychological barriers to quit attempts are cravings for the substance or activity from which people are trying to abstain. Elaborated Intrusion theory (Kavanagh, Andrade & May, 2005) explains cravings as cognitive-emotional states in which external or internal cues trigger intrusive thoughts (I need a drink) that are then elaborated, generating embodied images of the desired substance. These images are rich in sensory detail (the appearance, smell and taste of a drink), simulating the desired experience and conveying the pleasure or relief of the real thing. Being proximal and concrete, these highly vivid images dominate experience and drive out the intention to abstain. I shall review evidence from laboratory and field studies testing EI theory, and present some preliminary data on a novel motivational intervention called Functional Imagery Training, or FIT. The focus of FIT is on making the imagery associated with succeeding in a quit attempt richer and more concrete, so that it can compete with the shorter term temptations, and help people to withstand them.
25/11/2015; 4pm; Link3 workspace area
Psychology Research Seminar hosted by: Andy Willis
Dr. Kyle Wedgwood (University of Exeter)
At the single cell level, neurons typically exhibit an all-or-nothing response, dependent on the summation of input currents they receive from the rest of the network. Due to far reaching processes, neurons can form connections with distant parts of the network, allowing for rapid communication across long distances.
Certain neural systems show computation through patterned activity: persistent localised activity, in the form of bumps has been linked to working memory, whilst the propagation of activity in the form of waves has been associated with binocular rivalry tasks.
The assumption of infinitely slow synapses allows for the replacement of firing patterns with firing rates, resulting in a neural field model that is amenable to perturbative analysis. This description of the network averages out fluctuations in both space and time, ignoring these small scale effects. Our aim is to perform analysis on a network that retains these small scale effects, but whose large scale effects can be predicted in an analogous way to neural field models.
We present analysis of a network of minimal three-state neurons whose transitions are probabilistic. By taking appropriate limits, we demonstrate the existence and compute stability of spatiotemporal patterns of activity across the network. We then go on to show how coarse-grained analysis can be used to construct bifurcation diagrams for the network when these limits are relaxed and show how these can be used to reduce the complexity of the dynamics.
27/11/2015; 2pm; Link3 workspace area
CRNS Seminar hosted by: Martin F. Stoelen
Christian Fullgrabe (MRC Institute of Hearing Research, Nottingham)
Anecdotal evidence and experimental investigations indicate that older people experience increased speech-perception difficulties, especially in noisy environments. Since peripheral hearing sensitivity declines with age, lower speech intelligibility can often be explained by a reduction in audibility. However, aided speech-perception in hearing-impaired listeners frequently falls short of the performance level that would be expected based on the audibility of the speech signal. Given that many of these listeners are older, poor performance may be caused by age-related changes in supra-threshold auditory and/or cognitive processes that are not captured by the standard clinical assessment - the audiogram. The presentation will discuss experimental evidence obtained from clinically normal-hearing adults showing that auditory temporal processing, cognition (e.g. processing speed, attention, memory), and speech-in-noise processing (from phoneme identification to paragraph comprehension) are indeed linked and, independently of hearing loss, decline across the adult lifespan. These findings highlight the need to take into account these audibility-unrelated factors in the prediction and rehabilitation of speech processing across adulthood.
2/12/2015; 4pm; Link3 workspace area
Psychology Research Seminar hosted by: Sue Denham
Helen Haste (University of Bath)
Stories are the shared memories and aspirations through which we make meaning. They give us explanations about cause and effect, and about what is important to attend to in the past. They position us in relation to others and other groups. Stories both shape and reflect our identity, and they fuel our efficacious engagement with social issues. Attitude measurement, the "gold standard" of social research, can at best only capture the superficial level of beliefs and especially of motives. Drawing on data from China and South Africa, I argue that we should be seeking explanations of civic and social action and civic identities in the narratives that are central to people's identities
9/12/2015; 4pm; Link3 workspace area
Psychology Research Seminar hosted by: Sylvia Terbeck
Serena Ivaldi (INRIA, France)
The interaction between the human and the robot is a complex mechanism involving the exchange of numerous signals. The analysis and control of social and physical signals is fundamental to enable natural interaction between the partners and learning mechanisms.
I will present some experiments that we performed with the humanoid robot iCub interacting with people for
11/12/2015; 2pm; Link3 workspace area
CRNS Seminar hosted by: Samantha Adams and Martin F. Stoelen
In this short subseries of seminars ‘Ephemeral Affections’ point to a connection of ephemerality and movement between the subject matters of the idea of a machine (Jane Hutchinson), the infra-mince (Edith Doove) and of touch (Abigail Jackson).
Situated within art history and media archaeology the starting point for this specific seminar is a small bubble spirit level that has been used in a curator’s practice. As a continuation of the earlier seminar ‘The Opaque Lens’ by Edith Doove (February 2015) and the measurement of adequate behaviour in the ‘motor man machine’ as discussed by Jane Hutchinson, the bubble spirit level leads to an investigation of the general idea of equilibrium, measurement and measuring, the critical attitude towards it at the beginning of the 20th century (Poincaré), an implicate order (Bohm) and a necessary ‘mis-take’ (Duchamp).
The seminar will include a practical exercise in constructing a subjective standard.
References:
Bohm, D. (1994). Thought as a system. London: Routledge.
- (1996). On Creativity. New York: Routledge.
Molderings, H. (1991). Objects of Modern Scepticism. In: T. De Duve, ed., The Definitely Unfinished Marcel Duchamp. Cambridge, Massachusetts: MIT Press, pp.243-65.
Poincaré, H. (1905). Science and hypothesis. London et al: Walter Scott Publishing Co. Available on www.gutenberg.org
16/12/2015; 1pm; Link3 workspace area
TransTechnology Seminar "Objects of Affect and Affection" hosted by: Hannah Drayson and Michael Punt
Sara Invitto (University of Salento, Italy)
A haptic interface is a device that allows you to manoeuvre a robot, real or virtual, and receive feedback of tactile sensations. The user, in a haptic interface, is able to produce and to use motor actions such as physically manipulating the interface, which in turn displays and stimulates tactile kinaesthetic sensory information. Specifically, when subjects interact with objects in the physical world, they implement intentional schemes enabling them to place and represent the effects of their actions. Users in virtual reality (VR), especially if accustomed to interacting with digitalised 3D environments, must adapt to these schemes in the simulated world. The subjects are able to do so through continuous work of creating correspondences between the natural world (in which they continue to move and have most of their sensory information) and the simulated world (in which they see the effect of their intentional actions). VR allows for a new type of human-computer interaction, using the latter not as an expert system or a calculation tool, but as an artefact capable of conveying perceptual/motor processes that the subject produces. The purpose of this seminar will be to illustrate a new view extending beyond the interfaces of VR in neurocognitive protocols, highlighting in particular the relationship between haptic interaction in augmented or virtual reality, intentional motor and sensory processes and motor imagery. I will analyse the work through an ERP analysis, and compare motor imagery training with haptic training of augmented reality, as well as handling training of real objects with grasping affordances.
It will also be analysed for a job comparing the manipulation training in augmented reality and the manipulation of 3D models with difficult categorisation (planktonic elements). Finally, a further contribution to the seminar will be a description of a model (under construction in the Laboratory Nabidit - CNR Nano - Lab for Neuroscience Unisalento and Sensichip) of a haptic effector sending signals of motion, flexion of the fingers, skin conductance, as well as sending stimuli about the grain and texture of the surface to the subject.
22/01/2016; 2pm; Link3 workspace area
CRNS Seminar hosted by: Angelo Cangelosi and Martin F. Stoelen
Using a Geyer 8mm film splicer as starting point we will investigate the practice of cinema as an act of cutting and joining together, a co-produced work initiated by filmmakers but completed by film audiences. By dissecting the relative quality of cuts and joins and the tools used to make them we can trace their influence in the cinematic apparatus from a material to a perceptual level and extend the investigation into the broader question of the nature of the interframe space. The knives under discussion will therefore also include the rotating blades of the shutter in the analogue film projector. Is there an overlooked component of the affective potential of cinema quietly resident between the frames of film? Wherefore such a delicate commodity in a cinema of digitized systems of editing and projection, manufacture and distribution?
Murch, W. (2001). In the blink of an eye: A perspective on film editing. Silman-James Press.
Stewart, G. (1998). Cinecriture: Modernism's flicker effect. New Literary History, 29(4), 727-768.
17/02/2016; 1pm; Link3 workspace area
TransTechnology Seminar "Objects of Affect and Affection" hosted by: Hannah Drayson and Michael Punt
Khiet Truong (University of Twente)
When we interact with each other, not only the content of the words matter (what we say), but also the manner in which these words are spoken matter (how we speak), as well as the body language. Non-verbal behavior plays a key role in communicating affective and social information in human-human interaction. With the increasing acceptance of technology in our daily lives, such as virtual agents and robots, the need for developing technology that can sense and interpret human affect and social signals increases as well.
In this talk, I will discuss what kinds of non-verbal behaviors (mainly in speech communication, for example, laughter, backchannel behavior) can be important in human-agent interaction. How can we model these non-verbal behaviors for affective and social human-agent interaction? Our starting point is human-human interaction: how do humans display affect and social behavior? I will present some results of our studies on analysis of affective and social signals in speech and show how these results can inform the development of socially interactive agents.
Khiet Truong is an assistant professor with the Human Media Interaction (HMI) group at the University of Twente, Enschede, the Netherlands. She works in the fields of affective computing and social signal processing. Her main research interests lie in analysing and understanding emotionally expressive and social behaviors in interactions between humans, as well as in interactions between humans and virtual/physical agents (robots). Using this understanding, her aim is to develop socially intelligent and affective technology. She is particularly interested in paralinguistics: how do people talk in interaction and how can we develop technology that can automatically analyse and interpret the way people talk? At HMI, she is and was involved in several large EU-projects such as SQUIRREL, TERESA, and SSPNet.
22/02/2016; 12pm; Link3 workspace area
CogNovo Research Seminar hosted by: Ilaria Torre
Caroline Rowland (University of Liverpool)
Research on language development, particularly the development of grammar, has traditionally focused on debating the extent to which language learning depends on innate knowledge or environmental support. On the one hand, many studies, mainly on speech production (e.g. Pine et al., 1998), have suggested that children start out with pockets of knowledge based round an inventory of item-based frames. This evidence supports an approach that sees grammar development as a gradual process of abstraction across specific instances in the child's input. On the other hand, a different body of work, mainly on language comprehension, suggests that children use abstract grammatical categories from the earliest age tested (e.g. Gertner et al., 2006). This evidence supports an approach that proposes innate syntactic, semantic or conceptual knowledge at the core of grammar acquisition, and which predicts more rapid learning.
However, recent work suggests that this is a false dichotomy; children and adults have both abstract knowledge and knowledge centred around lexical items at all stages of development. Thus, the traditional approaches are breaking down. What is replacing them is a focus on explanatory models designed to answer a different question: "How do the child's learning mechanisms exploit information in the environment to build mature linguistic knowledge?" In this talk I use recent work from our lab to demonstrate what this approach has taught us so far about grammar acquisition, focussing on work that demonstrates what kind of learning mechanism best explains developmental differences in structural priming. I show how this new approach requires that we factor into our models the mechanisms underlying language processing, since the results of all studies on children's language development reflect not only children's knowledge of their language, but also the processing constraints that operate when we produce or comprehend language.
24/02/2016; 4pm; Link3 workspace area
Psychology Research Seminar hosted by: Caroline Floccia
Fumihide Tanaka (University of Tsukuba, Japan)
n this talk, I’ll introduce our past 10 year’s activities related to educational robots. The topics include
Bio: Ph.D. at Tokyo Institute of Technology (2003). R&D for entertainment robots at Sony Corporation, Digital Creatures Laboratory (2003-2008). Started educational robot researches at University of California, San Diego (2004-2007). Moved back to academia since 2008, The University of Tokyo (-2014). Best Paper Award at IEEE RO-MAN 2005, etc.
29/02/2016; 1pm; Link3 workspace area
CRNS Seminar hosted by: Tony Belpaeme and Martin F. Stoelen
Neil Ferguson (Liverpool Hope University)
This presentation explores the processes involved in leaving social movements or disengaging from terrorist activities by providing an analysis of the Ulster Volunteer Force (UVF) and Red Hand Commando (RHC) transformation away from politically motivated violence towards a civilian non-military role. Interpretative phenomenological analysis was employed to gain an understanding of participant accounts of leaving violence behind and disengaging from terrorism. Analysis of the interview transcripts revealed the interplay of individual, organization and societal level processes in incentivizing and obstructing disengagement from politically motivated violence. The findings resonate with other case studies exploring the processes involved in disengagement from political violence among other terror groupings across the globe. The results are discussed in relation to a number of topics, including the implementation DDR in post-conflict societies, the dynamic role of collective identity in the engagement in and disengagement from politically motivated violence and the role of prison in shaping disengagement from politically motivated violence. Keywords: disengagement, de-radicalization, terrorism, Northern Ireland, political violence.
02/03/2016; 4pm; Link3 workspace area
Psychology Research Seminar hosted by: Sylvia Terbeck
Reinout Wiers (Universiteit van Amsterdam)
Dual process models have described addiction as a combination of relatively strong bottom-up cue-related neurocognitive processes and relatively weak top-down cognitive control processes. In line with this perspective, we found across several studies a larger impact of memory associations and approach tendencies on behaviour in adolescents with relatively weak cognitive. Dual-process models have recently come under fire, but we think they can still be useful at a descriptive psychological level, while more work should be done to illuminate the underlying neurocognitive mechanisms. Moreover, dual process models inspired new interventions aimed at changing relatively automatic processes in addiction, varieties of Cognitive Bias Modification (CBM) paradigms. I will present work on attentional re-training in alcoholism and on approach-bias re-training which have yielded clinically relevant results. I will also present some recent studies concerning online applications of CBM and on the underlying neurocognitive mechanisms in these training studies.
09/03/2016; 4pm; Link3 workspace area
Psychology Research Seminar hosted by: Jon May
Jelena Havelka (University of Leeds)
It has recently been demonstrated that immediate memory for digits is superior when items are presented in a meaningful 'keypad' spatial configuration. This phenomenon, termed 'visuospatial bootstrapping', involves the integration of verbal and spatial information in working memory via stored knowledge in long-term memory. We have recently explored the basis of this effect experimentally using dual-task manipulations, with outcomes indicating contributions to verbal-spatial binding from spatial working memory and modality-general storage (possibly within the episodic buffer). We have also examined the extent to which the effect emerges in different population groups, including children of different ages, healthy older adults, and individuals with mild cognitive impairment. An overview of this recent work will be provided, along with a consideration of current and future directions.
16/03/2016; 4pm; Link3 workspace area
Psychology Research Seminar hosted by: Natalie Wyer
In 1979, Gunpei Yokoi, traveling on a bullet train, saw a bored businessman playing with an LCD calculator by pressing the buttons. Yokoi then thought of an idea for a watch that doubled as a miniature game machine for killing time. Nintendo’s Game & Watch series of portable devices that played electronic games may not have been the first to do, but it helped influence the direction the company took as it transitioned into a defining element of the videogames medium. The handheld videogame system is just one means of interacting within a digital space via a physical device, but it is the handheld that provides a greater element of freedom and one that transcends obstacles that can interfere with other means of play. That is why the Game & Watch and subsequent handheld systems are so important for the process of play and how it affords its unique impact via its specific technological attributes. This seminar will attempt to explore how the way in which players interact with these physical devices and how over just a short period of time the methods of play have changed and how upcoming technological changes might impact upon this.
23/03/2016; 1pm; Link3 workspace area
TransTechnology Seminar "Objects of Affect and Affection" hosted by: Hannah Drayson and Michael Punt
Ian Apperly (University of Birmingham)
A growing literature on perspective-taking paints a complex picture. Perspective-taking may be spatial or social; automatic or controlled; and clearly depends on multiple cognitive mechanisms. I will describe some recent results from adults and children that suggest there is order in this chaos. One reason why we should care about this because it provides a powerful framework for investigating individual differences in healthy and pathological perspective-taking.
23/03/2016; 4pm; Link3 workspace area
Psychology Research Seminar hosted by: Marina Wimmer
With a focus on MRI technology we will open a discussion regarding the impact of this scanning technique as an observational method in medicine. Highlighting its implications for diagnosis, social conditions and neuroscience as a whole. Specifically, we will unpack rhetorics and visual strategies of SCAN (social, cognitive and affective Neuroscience) disciplines using MRI technologies to create ways of embedding their knowledge into a bigger dispositif of academic and cultural knowledge production. We will consider whether forms of brain modelling, such as MRI, can offer a suitable comparison to our comprehension of the plasticity of the brain’s structure and reflect on the further application of this type of data by looking at critical neuroscience through the perspectives of differing mediums and disciplines.
Major texts:
13/04/2016; 1pm; Link3 workspace area
TransTechnology Seminar "Objects of Affect and Affection" hosted by: Hannah Drayson and Michael Punt
Nicola Byrom (St John's College Oxford)
04/05/2016; 4pm; Link3 workspace area
Psychology Research Seminar hosted by: Andy Willis
Jonathan Fieldsend (University of Exeter)
Evolutionary computation (EC) techniques are now extensively used when attempting to discover the optimal or near optimal parametrisation for problems with complex function transformation from parameters (design variables) to objective(s). Almost all optimisation procedures search the parameter space by evaluating the objectives for a given parametrisation before proposing a new, hopefully better, parametrisation. It is generally assumed that repeated evaluation of the objectives for a single parametrisation yields the same objective values. However, in many problems there is additional uncertainty in the veracity of the results obtained from the system model. Clear examples arise in “embodied” optimisation, where measurement error or stochastic elements in a physical system leads to different results for repeated evaluations at the same parameter values, or when the objectives are derived from Monte Carlo simulations or data-driven systems.
Work in the scalar optimisation EC community to tackle these types of problem has indicated that elitist techniques often prove fragile, as there is no longer the guarantee that fitness of the elite solution improves as an optimiser progresses. As most modern multi-objective evolutionary algorithms (MOEAs) rely heavily on elitism, this especially of concern when optimising uncertain multi-objective problems.
This talk will focus on the problem of uncertain evaluation, where the uncertainty can be modelled as additive noise in objective space. The rolling tide evolutionary algorithm (RTEA) will be presented, which progressively improves the accuracy of its estimated Pareto set, whilst simultaneously driving the front towards the true Pareto front. It can cope with noise whose characteristics change as a function of location (both design and objective), or which alter during the course of an optimisation. An effective data structures required to make the approach computationally feasible will also be described, which also has potential uses in dynamic optimisation problems. The talk will conclude with current work exploring issues with the probability of dominance which is currently employed in the domain, and also provide some examples of visualisation of multi-objective data sets.
06/05/2016; 2pm; Link3 workspace area
CRNS Seminar hosted by: Angelo Cangelosi, Guido Bugmann, and Liz Stuart
Steve Davis (University of Salford)
The future of manufacturing will likely see robots and humans working collaboratively and we are already seeing robots operating in new application areas. For safety reasons industrial robots have historically been kept away from humans with cages often separating the two, however, this makes interaction and collaboration impossible. There is therefore a need to develop methods to allow robots and humans to operate safely in a shared workspace. The two approaches to solving this challenge are through software or novel designs of robot hardware. This seminar will introduce new robot designs being developed which include compliance, variable stiffness and physical softness to allow safer human interaction in the manufacturing sector and beyond.
09/05/2016; 2pm; Link3 workspace area
CRNS Seminar hosted by: Martin Stoelen
Nick Peres and Darren Woodall
"They told me to sit on the bed and relax. Then they connected this machine to me that started to make all sorts of noises and beeping. This was fine when the nurse was with me, but after a short while I was left on my own with a beeping machine and no one to tell me what it was doing. Staring at the numbers on it I kept thinking, is everything normal? Is it meant to be doing that? All the while the nurse and others kept walking past where I was, just occasionally glancing in at it. I guess everything must have been ok" (Extract from a patient interview, February 2016).
By focusing on the vital signs machine and its role as a primary tool for capturing patient observations, this seminar will open for discussion how the vital signs machine could be interpreted as a gateway to the patient condition. Highlighting its interface of digits and bleeps as a shifting point of interaction between clinician and patient, we will attempt to unpack whether this current mechanism for capturing data has replaced a more personalised pre technology approach to this relationship and in the process submerged an empathetic connection between doctor and the patient.
We will explore, by using a vital signs machine during the seminar, how it feels to the individual to have personal data such as blood pressure, pulse, oxygen levels and temperature displayed as an interface of numbers and bleeps to the group. This exercise will aim to show the more exposed way in which this information can be ‘glanced’ with such an interface, as well as open some discussion on the trajectory of vital signs recording and the use of such ‘exposed’ data in shaping a future digital health service.
Do you prefer to be treated or do you want to feel better? Leading on from the above, the seminar will explore whether there may be advantage in having distance of relationship between doctor and patient afforded by such a digital interface, and if so at what point does the notion of the affectionate clinician become a genre?
11/05/2016; 1pm; Link3 workspace area
TransTechnology Seminar "Objects of Affect and Affection" hosted by: Hannah Drayson and Michael Punt
The last 20 years have seen an explosion of interest in the self within cognitive science. However, research on this topic has often been disjointed with researchers from cognitive neuroscience emphasising the importance of a bodily form of self which is formed by the integration of sensory inputs and motor outputs while researchers from the social sciences have tended to view the self as an abstract conceptual structure. In this talk I will present a series of studies which investigated whether bodily and conceptual forms of self-representation interact with one another and how this affected our perceptions of other people.
I will first present a series of studies which investigated the effect of skin colour on body ownership and found that experiencing body ownership over a hand with the skin colour of a racial out-group led to more positive implicit attitudes towards members of that racial out-group and modulated their empathic motor resonance to painful stimuli on the hand of that out-group member. I will go on to discuss a second series of studies that examined the relationship between trust and body representation using economic games and fMRI.
11/05/2016; 4pm; Link3 workspace area
Psychology Research Seminar hosted by: Sylvia Terbeck
Nicolas Pugeault (University of Exeter)
Driving a car in an urban setting is an extremely difficult problem, incorporating a large number of complex visual tasks; yet, this problem is solved daily by most adults with little apparent effort. For robots this remains a challenge: despite progress in computer vision over the last decades, artificial vision systems remain far from human vision in performance, robustness and speed. As a consequence, current prototypes of self-driving cars rely on a wide variety of sensors to palliate the limitations of their visual perception. How is it possible that such a complex task requires seemingly so little attention - to the point that inattention is cited as a leading cause of road accidents?
This talk will present recent research showing that a large proportion of a driver's actions can be explained by a simple pre-attentive model, and that such a model is not only accurate enough to steer a car on a road, but can also predict the driver's actions up to one full second before he starts turning the wheel.
13/05/2016; 2pm; Link3 workspace area
CRNS Seminar hosted by: Angelo Cangelosi and Martin Stoelen
Marc Goodfellow (University of Exeter)
Epilepsy is characterised by the repeated occurrence of seizures, which are periods of pathological brain activity that arise spontaneously from a predominantly healthy functional state. Since the goal of epilepsy treatment is to abolish or reduce the tendency of the brain to transition into seizures (its ictogenicity), it is important to better understand these transitions, and how we might interact with the brain to abate them. However, seizure dynamics emerge in, and affect, large-scale brain networks, and the network paradigm for ictogenesis introduces unfamiliar challenges and new opportunities to understand epilepsy.
In this talk I will introduce a mathematical model-based approach to quantify ictogenicity in brain networks. I will demonstrate how this approach can be used to quantify differences in brain networks between patients with generalised epilepsies and healthy controls. I will also describe how we can extend this approach to quantify the contribution of each component of a network to seizure generation. This quantification is based upon the effect that a treatment-specific perturbation has on network ictogenicity. Using exemplar networks I will explore how the apparent ictogenicity of nodes can vary according to network structure and the presence or absence of "pathological" nodes (seizure foci). I will explain how this approach can potentially provide an insightful and principled way to interpret and describe generalised or focal seizure dynamics, and may enhance our strategies for the classification and treatment of epilepsies.
20/05/2016; 2pm; Link3 workspace area
CRNS Seminar hosted by: Martin Stoelen
Jie Ma (China Research Laboratory at IBM,)
The talk demonstrates some recent projects Dr Ma has been doing in Robotics and IoTs in UK and IBM China. In particular the talk highlights a cognitive biped locomotion method Ma proposed for humanoid robots that can teach a robot learn to walk by itself like a human baby using reinforcement learning methods. The talk also includes some recent cognitive learning projects in manufacturing and industry 4.0 in IBM China and the introduction of cognitive computing platform of IBM in IoTs and Robotics domains.
Dr Jie Ma is a research scientist in the China Research Laboratory at IBM, with a specialisation in cognitive computing, machine learning, IoTs and Robotics. After receiving his doctoral degree from Department of Computer Science, University of Oxford in 2011, Dr Ma became a research fellow in the Cognitive Robotics Research Centre (CRRC) at University of Wales. His research involved cognitive learning algorithms for teams of agents in dynamic environments. Applications include strategy learning, cooperative planning, autonomous decision-making and biped locomotion. His recent publications include Human Robot Interaction using robot skin, Policy Search Planning in multi-agent cooperation and QWalking, a model-free biped locomotion learning approach for humanoid robots. His work is based on physical robots such as iCub, NAO and Kondo and other IoT devices. Dr Ma was also a member of organising and technical committees in some recent Robot World Cup (RoboCup) competitions.
27/05/2016; 2pm; Link3 workspace area
CRNS Seminar hosted by: Torbjørn S. Dahl
Daniel Braun ( Max Planck Institute for Biological Cybernetics)
Recent advances in movement neuroscience suggest that sensorimotor control can be considered as a continuous decision-making process in complex environments in which uncertainty and task variability play a key role. Leading theories of motor control assume that the motor system learns probabilistic models and that motor behavior can be explained as the optimization of payoff or cost criteria under the expectation of these models. Here we discuss how the motor system exploits task variability to build up efficient models and then discuss evidence that humans deviate from Bayes optimal behavior in their movements, because they exhibit effects of model uncertainty. Furthermore, we discuss in how far model uncertainty can be considered as a special case of a general information-processing and decision-making framework inspired by statistical physics and thermodynamics.
02/06/2016; 2pm; Link3 workspace area
CRNS Seminar hosted by: Ian Howard
The representation of the "distilled moment" or "decisive moment", clarifies some things whilst it obscures others; what it obscures is the distinction between attention and consciousness (awareness). The close relationship between attention and consciousness has led many scholars to conflate these processes whilst everyday speech and literature often resists a clear definition. This seminar will provide examples that show consciousness and attention as two discrete components that need not occur together. This, in turn, has implications for the way we think about photographic practices and the weight of the collaboration between the photographer, the camera and the world revealed in certain incidental outputs. This seminar will set up a problematic that aims to untangle the tight relationship between attention and consciousness through a range of photographic examples where there is a dominance of one or the other in the representation of a seemingly fleeting moment.
Suggested Reading:
15/06/2016; 1pm; Link3 workspace area
TransTechnology Seminar "Objects of Affect and Affection" hosted by: Hannah Drayson and Michael Punt
For more up-to-date information and titles/abstracts please visit the relevant research group websites:
