A5038997077- Robot Manipulation and Learning
- Action Observation and Synchronization
- Motor Control and Adaptation
- Visual perception and processing mechanisms
- Neural dynamics and brain function
- EEG and Brain-Computer Interfaces
- Reinforcement Learning in Robotics
- Gaze Tracking and Assistive Technology
- Soft Robotics and Applications
- Robotic Locomotion and Control
- Human Pose and Action Recognition
- Tactile and Sensory Interactions
- Face Recognition and Perception
- Child and Animal Learning Development
- Modular Robots and Swarm Intelligence
- Social Robot Interaction and HRI
- Vestibular and auditory disorders
- Muscle activation and electromyography studies
- Neural and Behavioral Psychology Studies
- Context-Aware Activity Recognition Systems
- Robotic Mechanisms and Dynamics
- Anomaly Detection Techniques and Applications
- Robotics and Automated Systems
- Advanced Vision and Imaging
- Psychology of Moral and Emotional Judgment
Middlesex University
2017-2023
University of Leeds
2014-2016
Imperial College London
2011-2013
Universitat Jaume I
2003-2011
University of Edinburgh
2004-2005
In human-human interactions, corepresenting a partner's actions is crucial to successfully adjust and coordinate with others. Current research suggests that action corepresentation restricted interactions between human agents facilitating social interaction conspecifics. this study, we investigated whether corepresentation, as measured by the Simon effect (SSE), present when share task real humanoid robot. Further, tested believed humanness of robot's functional principle modulates extent...
Thanks to the efforts of robotics and autonomous systems community, robots are becoming ever more capable. There is also an increasing demand from end-users for service that can operate in real environments extended periods. In STRANDS project we tackling this head-on by integrating state-of-the-art artificial intelligence research into mobile robots, deploying these long-term installations security care environments. Over four deployments, our have been operational a combined duration 104...
Reaching and grasping an object is action that can be performed in light, as well darkness. Area V6A a visuomotor area of the medial posterior parietal cortex involved control reaching movements. It contains neurons modulated by passive somatosensory visual stimulations. In present work we analyze effect feedback on activity neurons. Three macaques were trained to execute movements two conditions: darkness, where only target was visible, full monkey also saw its own moving arm environment....
One of the defining characteristics human cognition is our outstanding capacity to cooperate. A central requirement for cooperation ability establish a "shared plan"-which defines interlaced actions two cooperating agents-in real time, and even negotiate this shared plan during its execution. In current research we identify requirements cooperation, extending earlier work in area. These include using spoken language, learn new component within that plan, based on visual observation...
Primates often perform coordinated eye and arm movements, contextually fixating reaching towards nearby objects. This combination of looking to the same target is used by infants establish an implicit visuomotor representation peripersonal space, useful for both oculomotor motor control. In this work, taking inspiration from such behavior primate mechanisms, a shared sensorimotor map environment, built on radial basis function framework, configured trained control movements. Computational...
Autonomous manipulation is a key issue for humanoid robot. Here, we are interested in vision-based grasping behavior so that the robot can deal with previously unknown objects real time and an intelligent manner. Starting from number of feasible candidate grasps, focus on problem predicting their reliability using knowledge acquired previous experiences. A set visual features which take into account physical properties affect stability grasp defined. obtains its experience by repeating large...
This paper presents and analyzes 12 quality measures that characterize robotic grips according to their stability reliability. The are designed assess three-finger of two-dimensional parts performed in a real environment, taking into account both theoretical aspects unavoidable uncertainties grasping action. They build on the existing literature physical mechanical considerations. constitute feature space pattern recognition methods can use order classify quality. Six depend actual finger...
This paper presents and analyses ten criteria that assess the quality of a set three-finger grips suitable for dextrous manipulation on real 2D parts. The candidate hand configurations is result previous process grasp generation from object image. proposed include six depend actual finger configuration gripper. kinematics Barrett has been used. are merged to give global value can be used select best grip execute. Experimental results tests stability effect parameter variation.
Previous kinematic research suggests that visually guided grasping employs an accurate real-time control system in the dorsal stream, whereas delayed relies on less stored information derived by perceptual ventral stream. We explored these ideas two experiments combining and with auditory tasks involving perception-based imagery semantic memory. In both experiments, participants were cued to grasp three-dimensional objects of varying sizes. During trials, visible during interval between cue...
Neurodecoders have been developed by researchers mostly to control neuroprosthetic devices, but also shed new light on neural functions. In this study, we show that signals representing grip configurations can be reliably decoded from data acquired area V6A of the monkey medial posterior parietal cortex. Two Macaca fascicularis monkeys were trained perform an instructed-delay reach-to-grasp task in dark and toward objects different shapes. Population activity was extracted at various time...
Exploratory gaze movements are fundamental for gathering the most relevant information regarding partner during social interactions. Inspired by cognitive mechanisms underlying human behaviour, we have designed and implemented a system dynamic attention allocation which is able to actively control visual action recognition task exploiting its own execution predictions. Our humanoid robot able, observation of partner's reaching movement, contextually estimate goal position hand location in...
Reaching a target object in an unknown and unstructured environment is easily performed by human beings. However, designing humanoid robot that executes the same task requires implementation of complex abilities, such as identifying visual field, estimating its spatial location, precisely driving motors arm to reach it. While research usually tackles development abilities singularly, this work we integrate number computational models into unified framework, demonstrate torso feasibility...
This paper presents an approach to facial expression recognition based on the theory of liquid computing. Up date, no emotion systems spiking neural networks exist, and our work is first attempt in this direction. We investigated pattern ability Liquid State Machines various models, such as integrate-and-fire, resonate-and-fire, FitzHugh-Nagumo, Morris-Lecal, Hindmarsh-Rose Izhikevich's models. No single Machine provided particularly good results, but a global classifier we defined merging...
The properties of separation ability and computational efficiency liquid state machines depend on the neural model employed connection density in column. A simple part mammalians visual system consisting one hypercolumn was examined. Such a stimulated by two different input patterns, Euclidean distance, as well partial global entropy column responses were calculated. Interesting insights could be drawn regarding models used hypercolumn, effect information representation capability system.
Saccades are fast eye movements that allow humans and robots to bring the visual target in center of field. open loop with respect vision system, thus their execution require a precise knowledge internal model oculomotor system. In this work, we modeled saccade control, taking inspiration from recurrent loops between cerebellum brainstem. model, brainstem acts as fixed-inverse while an adaptive element learns The filter is implemented using state-of-the-art neural network, called I-SSGPR....
Action observation and execution share overlapping neural substrates, so that simultaneous activation by modulates motor performance. Previous literature on simple prehension tasks has revealed influence can be two-sided: facilitation for observed performed congruent actions interference incongruent actions. But little is known of the specific modulations performance in complex forms interaction. Is it possible very same movement lead either to or effects a temporally executed action,...
Localized neuropathic pain (LNP) is a subgroup of characterized by consistent and circumscribed area(s) maximum pain, associated with negative or positive sensory signs and/or spontaneous symptoms characteristic NP. Lidocaine medicated plasters (LMP) have shown to be effective in relief selective LNP syndromes.We collected data 130 patients our database syndromes who used LMP.Forty-one out (32%) were treated antiepileptics, antidepressants opioids without improvement intolerable adverse...
Activity classification from smart environment data is typically done employing ad hoc solutions customised to the particular dataset at hand. In this work we introduce a general purpose collection of features for recognising human activities across datasets different type, size and nature. The first experimental test our feature achieves state art results on well known datasets, provide importance analysis in order compare potential relevance activity datasets.
This paper deals with visually guided grasping of unmodeled objects for robots which exhibit an adaptive behavior based on their previous experiences. Nine features are proposed to characterize three-finger grasps. They computed from the object image and kinematics hand. Real experiments a humanoid robot Barrett hand carried out provide experimental data. data is employed by classification strategy, k-nearest neighbour estimation rule, predict reliability grasp configuration in terms five...
In human–human interactions, a consciously perceived high degree of self–other overlap is associated with higher integration the other person's actions into one's own cognitive representations. Here, we report data suggesting that this pattern does not hold for human–robot interactions. Participants performed social Simon task robot, and afterwards indicated using Inclusion Other in Self (IOS) scale. We found no overall correlation between effect (as an indirect measure overlap) IOS score...