Figure 1 - available via license: Creative Commons Attribution 4.0 International
Content may be subject to copyright.
![A model of human event memory proposed by Rubin [4,16] adapted from Rubin [4]. See text for further details.](publication/330737808/figure/fig1/AS:749607155937280@1555731681514/A-model-of-human-event-memory-proposed-by-Rubin-4-16-adapted-from-Rubin-4-See-text.png)
A model of human event memory proposed by Rubin [4,16] adapted from Rubin [4]. See text for further details.
Source publication
![Figure 1. A model of human event memory proposed by Rubin [4,16]...](publication/330737808/figure/fig1/AS:749607155937280@1555731681514/A-model-of-human-event-memory-proposed-by-Rubin-4-16-adapted-from-Rubin-4-See-text_Q320.jpg)
From neuroscience, brain imaging and the psychology of memory, we are beginning to assemble an integrated theory of the brain subsystems and pathways that allow the compression, storage and reconstruction of memories for past events and their use in contextualizing the present and reasoning about the future—mental time travel (MTT). Using computati...
Contexts in source publication
Context 1
... A framework for understanding human memory Figure 1 illustrates a theoretical framework proposed by Rubin [4,16] and grounded on data from experimental psychology concerning the properties and capabilities of human autobiographical memory, and from neuropsychology and neuroimaging concerning its likely neural substrates. These findings include differences in recollection and ratings of memory vividness [17 -19], under varying conditions of visual and auditory imagery and emotional significance, and evidence of the localization of different aspects of memory processing in multiple brain regions during recollection episodes, as indicated by studies of people with amnesia [20] and by patterns in fMRI activity [21,22]. ...
Context 2
... generation must allow both pattern completion-the filling out of a memory based on a fragment or a partial cue, and pattern separationthe ability to recall, as distinct patterns, events that share some of the same sensory properties. More specifically, and as described next, in implementing a version of the model shown in figure 1, we will view the operation of the sensory and multisensory memory systems as being analogous to learning processes that discover useful low-dimensional LV descriptions of high-dimensional data, and that operate bi-directionally both to encode high-dimensional stimulus patterns and to reconstruct such patterns from their low-dimensional description. ...
Context 3
... above results demonstrate that we are already making progress in our effort to provide iCub with some of the components of the memory system illustrated in figure 1; we conclude by briefly discussing how this system could be extended towards a more complete model of human autobiographical memory and MTT, and how such a system might be deployed in future social robots. ...
Citations
... To achieve this goal, a complex human-robot interaction (HRI) and control architecture have to be designed to comprehend different software components allowing for efficient and simultaneous execution of multiple tasks and for providing essential capabilities, such as storing past events [20], constructing models of others' actions, beliefs, and intentions [12], modelling the domain knowledge, selecting actions and behaviours, and planning [16]. For example, in the JAMES project [6], a bartender robot was able to engage participants in conversation by producing facial expressions and lip-synchronising speech. ...
The BRILLO (Bartending Robot for Interactive Long-Lasting Operations) project aims to create an autonomous robotic bartender that can interact with customers while accomplishing its bartending tasks. In such a scenario, people's novelty effect connected to the use of an attractive technology is destined to wear off and, consequently, negatively affects the success of the service robotics application. For this reason, providing personalised natural interaction while people access its services is fundamental for increasing users' engagement and, consequently, their loyalty. In this paper, we present the developed three-layer ROS architecture integrating a perception layer managing the processing of different social signals, a decision-making layer for handling multi-party interactions, and an execution layer controlling the behaviour of a complex robot composed of arms and a face. Finally, user modelling through a beliefs layer allows for personalized interaction. We also present the results of both people's interaction, experience and performances in a real user case. The user study involved 116 participants and showed that BRILLO is considered an easy-to-use and attractive system by the users.
... The formation and recall of memories in multiple sensory modalities have exploited the robustness of Bayesian Latent Variable Models [18], being able to produce meaningful representations of sensory data through non-linear dimensionality reduction and uncertainty quantification [4]. Prescott et al. [12] proposed the establishment of this family of models as Simple Synthetic Memories (SSMs). Active compression of high-dimensional data, creation of fantasy memories, and recovering of observations from latent spaces were explored for: face recognition, audition, action discrimination and tactile interactions. ...
It has been previously demonstrated in robots that the mimicking of functional characteristics of biologic memory can be beneficial for providing accurate learning and recognition in circumstances of social human-robot-interaction. The effective encoding of social and physical salient features has been demonstrated through the use of Bayesian Latent Variable Models as abstractions of memories (Simple Synthetic Memories). In this work, we explore the capabilities of formation and recall of tactile memories associated to the encoding of geometric and spatial qualities. Compression and pattern separation are evaluated against the use of raw data in a nearest neighbour regression model, obtaining a substantial improvement in accuracy for prediction of geometric properties of the stimulus. Additionally, pattern completion is assessed with the generation of ‘imagined touch’ streams of data showing similarities to real world tactile observations. The use of this model for tactile memories offers the potential for robustly perform sensorimotor tasks in which the sense of touch is involved.KeywordsTactile memoriesRobot touchLatent variable spaceTactile data generation
... In so doing, it can provide a stringent test of the completeness of the underlying theories, particularly in terms of the purported role of target brain subsystems in the real-time coordination of sensing with action (10)(11)(12)(13). Robotic models have, for instance, added to our understanding of the role of the spinal cord/brainstem in motor pattern generation (14,15); the cerebellum in predictive control (16)(17)(18)(19)(20)(21); the basal ganglia as a key substrate for action selection and reinforcement learning (22)(23)(24)(25); the hippocampus as an attractor network that supports memory storage and retrieval (26)(27)(28)(29)(30); and the cerebral cortex as a locus for self-organizing somatotopic maps (31), multisensory convergence (32), mental imagery (33), and metacognitive control (34). ...
... Image from Martin Pearson, Bristol Robotics Laboratory (with permission). (C) iCub, a humanoid robot widely used to model human perception, cognition, motor control, and social interaction (13,29,(31)(32)(33). Image from the University of Sheffield (with permission). ...
Robotics is increasingly seen as a useful test bed for computational models of the brain functional architecture underlying animal behavior. We provide an overview of past and current work, focusing on probabilistic and dynamical models, including approaches premised on the free energy principle, situating this endeavor in relation to evidence that the brain constitutes a layered control system. We argue that future neurorobotic models should integrate multiple neurobiological constraints and be hybrid in nature.
... Additionally, the marketing literature has acknowledged mental time travel as potentially relevant to the construct of confidence as confidence consists of past patterns that emerge in the cognitive processes of pre-production and re-production (Simintiras et al., 2014). Finally, recent studies have explored brain networks and mental time travel in relation to humanoid robots and AI to support a humanoid robot's multimodal memory system and ability to learn as well as recognize faces (Prescott et al., 2019), demonstrating the growing importance of the concept of mental time travel as a construct. ...
This research introduces the concept of mental time travel experience (MTTE) to the experiential marketing literature as a new type of experience that marketers can evoke to prompt consumers to mentally travel to a different time in one’s personal past or forwards in time to one’s personal future. Although research indicates that experiences are a cornerstone of building positive brand equity for firms, and that humans are rarely mentally in the present moment, there is a distinct gap regarding experiences that take place in consumers’ minds. We address this gap with the introduction and examination of MTTEs with two between-subjects studies with 1,879 participants. Results identify three factors important to the elicitation of MTTEs and mental time travel to the past as a key factor to influencing behavioral intentions. Finally, findings indicate that lower states of immersion have the propensity to strengthen mental time travel’s effect on behavioral intentions.
... In the former case, all knowledge and control of the robot are housed within its onboard computers, which have limited computational memory and processing capacity. These limitations impede the ability of certain social robots to execute advanced AI algorithms, such as facial recognition [11], navigation [12], Natural Language Processing (NLP) [13], behavior analysis [14], etc. [15]. For the latter case, constraints arise from the complexity of the programming, which may make it difficult for non-experts to interact and operate the robots [16]. ...
Social robots have the potential to revolutionize the way we interact with technology, providing a wide range of services and applications in various domains, such as healthcare, education, and entertainment. However, most existing social robotics platforms are operated based on embedded computers, which limits the robot’s capabilities to access advanced AI-based platforms available online and which are required for sophisticated physical human–robot interactions (such as Google Cloud AI, Microsoft Azure Machine Learning, IBM Watson, ChatGPT, etc.). In this research project, we introduce a cloud-based framework that utilizes the benefits of cloud computing and clustering to enhance the capabilities of social robots and overcome the limitations of current embedded platforms. The proposed framework was tested in different robots to assess the general feasibility of the solution, including a customized robot, “BuSaif”, and commercialized robots, “Husky”, “NAO”, and “Pepper”. Our findings suggest that the implementation of the proposed platform will result in more intelligent and autonomous social robots that can be utilized by a broader range of users, including those with less expertise. The present study introduces a novel methodology for augmenting the functionality of social robots, concurrently simplifying their utilization for non-experts. This approach has the potential to open up novel possibilities within the domain of social robotics.
... GPs also grow in complexity to suit the data being therefore robust to overfitting [22]. By explicitly representing uncertainty in the data they can guide exploratory procedures aimed at reducing uncertainty [19], and, as a means of representing the data, they can be more transparent in terms of inspecting the low-dimensional manifold that is acquired by the trained system. In the current paper, we compare GP model with a newer variant, Deep GPs, that exploits hierarchical composition to create a deep belief network based on Gaussian process mappings [6]. ...
Active touch sensing can benefit from the representation of uncertainty in order to guide sensing movements and to drive sensing strategies that operate to reduce uncertainty with respect to the task at hand. Here we explore learning approaches that can acquire task knowledge quickly and with relatively small datasets and with the potential to be exploited for active sensing in robots and as models of biological sensory systems. Specifically, we explore the utility of deep (hierarchical) Gaussian Process models (Deep GPs) that have shown promise as models of episodic memory processes due to their low-dimensionality (compactness), generative capability, and ability to explicitly represent uncertainty. Using data obtained in a robotic active touch task (contour following), we show that both single-layer and Deep GP models are capable of providing robust function approximations from tactile data to angle and sensor position, with Deep GPs showing some advantages in terms of accuracy and uncertainty quantification in angle discrimination.KeywordsActive touchDeep Gaussian processContour followingTactile sensing
... For instance, to be genuinely helpful, the robot must be able to recognize individuals consistently, perhaps remembering past encounters, and be able to monitor and anticipate the person's needs, at least to some degree. Few, if any social robots, are capable of this level of helpful behavior at present [30]. On ethical behaviorism grounds, we might consider that the robot's statement that it "cares" and "wants" to help as problematic to the extent that it raises expectations about its wider behavior that cannot be met, however, a future, more care-capable robot might more reasonably make such statements. ...
There is increasing interest in social robots as assistive technologies to support a wide range of potential user groups. Nevertheless, the widespread use of robots has been challenged in terms of their efficacy and benefits as well as the ethics of employing robots in social roles. For instance, it has been suggested that robots are incapable of being truly social and therefore that any use of social robots as assistive technology is intrinsically deceptive. This contribution addresses this controversy, building on a relational view of human-robot interaction, which asserts that sociality has less to do with the essential natures of the human and robot actors involved, and more to do with the patterns and consequences of their interaction. From this starting position we consider and explore four design principles for social robots and compare/contrast these with the view of design "transparency" that robots should behave to reveal their true machine nature.
... As an example of the second category, Prescott et al. (2019) included emotional signals in a neuroscience-inspired multimodal computational architecture for the autobiographical memory system, named the mental time travel model, to control the iCub robot. The model allows for retrieving past events, including their emotional associations, and projecting them into an imagined future by using the same system. ...
The current state of the art in cognitive robotics, covering the challenges of building AI-powered intelligent robots inspired by natural cognitive systems. A novel approach to building AI-powered intelligent robots takes inspiration from the way natural cognitive systems—in humans, animals, and biological systems—develop intelligence by exploiting the full power of interactions between body and brain, the physical and social environment in which they live, and phylogenetic, developmental, and learning dynamics. This volume reports on the current state of the art in cognitive robotics, offering the first comprehensive coverage of building robots inspired by natural cognitive systems.
Contributors first provide a systematic definition of cognitive robotics and a history of developments in the field. They describe in detail five main approaches: developmental, neuro, evolutionary, swarm, and soft robotics. They go on to consider methodologies and concepts, treating topics that include commonly used cognitive robotics platforms and robot simulators, biomimetic skin as an example of a hardware-based approach, machine-learning methods, and cognitive architecture. Finally, they cover the behavioral and cognitive capabilities of a variety of models, experiments, and applications, looking at issues that range from intrinsic motivation and perception to robot consciousness.
Cognitive Robotics is aimed at an interdisciplinary audience, balancing technical details and examples for the computational reader with theoretical and experimental findings for the empirical scientist.
... For instance, to be genuinely helpful, the robot must be able to recognize individuals consistently, perhaps remembering past encounters, and be able to monitor and anticipate the person's needs, at least to some degree. Few, if any social robots, are capable of this level of helpful behavior at present [30]. On ethical behaviorism grounds, we might consider that the robot's statement that it "cares" and "wants" to help as problematic to the extent that it raises expectations about its wider behavior that cannot be met, however, a future, more care-capable robot might more reasonably make such statements. ...
There is increasing interest in social robots as assistive technologies to support a wide range of potential user groups. Nevertheless, the widespread use of robots has been challenged in terms of their efficacy and benefits as well as the ethics of employing robots in social roles. For instance, it has been suggested that robots are incapable of being truly social and therefore that any use of social robots as assistive technology is intrinsically deceptive. This contribution addresses this controversy, building on a relational view of human-robot interaction, which asserts that sociality has less to do with the essential natures of the human and robot actors involved, and more to do with the patterns and consequences of their interaction. From this starting position we consider and explore four design principles for social robots and compare/contrast these with the view of design “transparency” that robots should behave to reveal their true machine nature.
... To achieve this goal, a complex human-robot interaction (HRI) and control architecture have to be designed that comprehend different software components allowing for efficient and simultaneous execution of multiple tasks and for providing essential capabilities, such as storing past events [10], constructing models of others' actions, beliefs, desires, and intentions [11], modelling the domain knowledge, selecting actions and behaviours, and planning [12]. For example, in the JAMES project [9], a bartender robot was able to engage participants in conversation producing facial expressions and lip-synchronising speech. ...
BRILLO (Bartending Robot for Interactive Long-Lasting Operations) project has the overall goal of creating an autonomous robotic bartender that can interact with customers while accomplishing its bartending tasks. In such a scenario, people's novelty effect connected to the use of an attractive technology is destined to wear off and, consequently, it negatively affects the success of the service robotics application. For this reason, providing personalised natural interaction while accessing its services is of paramount importance for increasing users' engagement and, consequently, their loyalty. In this paper, we present the developed three-layers ROS architecture integrating a perception layer managing the processing of different social signals, a decision-making layer for handling multi-party interactions, and an execution layer controlling the behaviour of a complex robot composed of arms and a face. Finally, user modelling through a beliefs layer allows for personalised interaction.
