Manuel González Bedia

Artificial Intelligence approaches for the generation and assessment of believable human-like behaviour in virtual characters

Having artificial agents to autonomously produce human-like behaviour is one of the most ambitious original goals of Artificial Intelligence (AI) and remains an open problem nowadays. The imitation game originally proposed by Turing constitute a very effective method to prove the indistinguishability of an artificial agent. The behaviour of an agent is said to be indistinguishable from that of a human when observers (the so-called judges in the Turing test) cannot tell apart humans and non-human agents. Different environments, testing protocols, scopes and problem domains can be established to develop limited versions or variants of the original Turing test. In this paper we use a specific version of the Turing test, based on the international BotPrize competition, built in a First-Person Shooter video game, where both human players and non-player characters interact in complex virtual environments. Based on our past experience both in the BotPrize competition and other robotics and computer game AI applications we have developed three new more advanced controllers for believable agents: two based on a combination of the CERA-CRANIUM and SOAR cognitive architectures and other based on ADANN, a system for the automatic evolution and adaptation of artificial neural networks. These two new agents have been put to the test jointly with CCBot3, the winner of BotPrize 2010 competition (Arrabales et al., 2012), and have showed a significant improvement in the humanness ratio. Additionally, we have confronted all these bots to both First-person believability assessment (BotPrize original judging protocol) and Third-person believability assessment, demonstrating that the active involvement of the judge has a great impact in the recognition of human-like behaviour.

The situated HKB model: how sensorimotor spatial coupling can alter oscillatory brain dynamics

Despite the increase of both dynamic and embodied/situated approaches in cognitive science, there is still little research on how coordination dynamics under a closed sensorimotor loop might induce qualitatively different patterns of neural oscillations compared to those found in isolated systems. We take as a departure point the Haken-Kelso-Bunz (HKB) model, a generic model for dynamic coordination between two oscillatory components, which has proven useful for a vast range of applications in cognitive science and whose dynamical properties are well understood. In order to explore the properties of this model under closed sensorimotor conditions we present what we call the situated HKB model: a robotic model that performs a gradient climbing task and whose “brain” is modeled by the HKB equation. We solve the differential equations that define the agent-environment coupling for increasing values of the agents sensitivity (sensor gain), finding different behavioral strategies. These results are compared with two different models: a decoupled HKB with no sensory input and a passively-coupled HKB that is also decoupled but receives a structured input generated by a situated agent. We can precisely quantify and qualitatively describe how the properties of the system, when studied in coupled conditions, radically change in a manner that cannot be deduced from the decoupled HKB models alone. We also present the notion of neurodynamic signature as the dynamic pattern that correlates with a specific behavior and we show how only a situated agent can display this signature compared to an agent that simply receives the exact same sensory input. To our knowledge, this is the first analytical solution of the HKB equation in a sensorimotor loop and qualitative and quantitative analytic comparison of spatially coupled vs. decoupled oscillatory controllers. Finally, we discuss the limitations and possible generalization of our model to contemporary neuroscience and philosophy of mind.

Quantifying Political Self-Organization in Social Media. Fractal patterns in the Spanish 15M movement on Twitter

The objective of this work is to better analyse and understand social self-organization in the context of social media and political activism. More specifically, we centre our analysis in the presence of fractal scaling in the form of 1/f noise in different Twitter communication networks related to the Spanish 15M movement. We show how quantitative indexes of brown, white and pink noise correlate with qualitatively different forms of social coordination of protests: rigidly organized protests (brown noise), reactive-spontaneous protests (white noise) and complex genuinely self-organized protests (pink noise). In addition, pink noise processes present correlations that reach much further in time, maintaining a dynamical coherence that last several days, and also show a balance between mean distance and clustering coefficient within the interaction network.

Self-Organized Criticality, Plasticity and Sensorimotor Coupling. Explorations with a Neurorobotic Model in a Behavioural Preference Task

During the last two decades, analysis of 1/ƒ noise in cognitive science has led to a considerable progress in the way we understand the organization of our mental life. However, there is still a lack of specific models providing explanations of how 1/ƒ noise is generated in coupled brain-body-environment systems, since existing models and experiments typically target either externally observable behaviour or isolated neuronal systems but do not address the interplay between neuronal mechanisms and sensorimotor dynamics. We present a conceptual model of a minimal neurorobotic agent solving a behavioural task that makes it possible to relate mechanistic (neurodynamic) and behavioural levels of description. The model consists of a simulated robot controlled by a network of Kuramoto oscillators with homeostatic plasticity and the ability to develop behavioural preferences mediated by sensorimotor patterns. With only three oscillators, this simple model displays self-organized criticality in the form of robust 1/ƒ noise and a wide multifractal spectrum. We show that the emergence of self-organized criticality and 1/ƒ noise in our model is the result of three simultaneous conditions: a) non-linear interaction dynamics capable of generating stable collective patterns, b) internal plastic mechanisms modulating the sensorimotor flows, and c) strong sensorimotor coupling with the environment that induces transient metastable neurodynamic regimes. We carry out a number of experiments to show that both synaptic plasticity and strong sensorimotor coupling play a necessary role, as constituents of self-organized criticality, in the generation of 1/ƒ noise. The experiments also shown to be useful to test the robustness of 1/ƒ scaling comparing the results of different techniques. We finally discuss the role of conceptual models as mediators between nomothetic and mechanistic models and how they can inform future experimental research where self-organized critically includes sensorimotor coupling among the essential interaction-dominant process giving rise to 1/ƒ noise.

Unreliable gut feelings can lead to correct decisions: the somatic marker hypothesis in non-linear decision chains.

Dual-process approaches of decision-making examine the interaction between affective/intuitive and deliberative processes underlying value judgment. From this perspective, decisions are supported by a combination of relatively explicit capabilities for abstract reasoning and relatively implicit evolved domain-general as well as learned domain-specific affective responses. One such approach, the somatic markers hypothesis (SMH), expresses these implicit processes as a system of evolved primary emotions supplemented by associations between affect and experience that accrue over lifetime, or somatic markers. In this view, somatic markers are useful only if their local capability to predict the value of an action is above a baseline equal to the predictive capability of the combined rational and primary emotional subsystems. We argue that decision-making has often been conceived of as a linear process: the effect of decision sequences is additive, local utility is cumulative, and there is no strong environmental feedback. This widespread assumption can have consequences for answering questions regarding the relative weight between the systems and their interaction within a cognitive architecture. We introduce a mathematical formalization of the SMH and study it in situations of dynamic, non-linear decision chains using a discrete-time stochastic model. We find, contrary to expectations, that decision-making events can interact non-additively with the environment in apparently paradoxical ways. We find that in non-lethal situations, primary emotions are represented globally over and above their local weight, showing a tendency for overcautiousness in situated decision chains. We also show that because they tend to counteract this trend, poorly attuned somatic markers that by themselves do not locally enhance decision-making, can still produce an overall positive effect. This result has developmental and evolutionary implications since, by promoting exploratory behavior, somatic markers would seem to be beneficial even at early stages when experiential attunement is poor. Although the model is formulated in terms of the SMH, the implications apply to dual systems theories in general since it makes minimal assumptions about the nature of the processes involved.

Modeling flocks with perceptual agents from a dynamicist perspective

Computational simulations of flocks and crowds have typically been processed by a set of logic or syntactic rules. In recent decades, a new generation of systems has emerged from dynamicist approaches in which the agents and the environment are treated as a pair of dynamical systems coupled informationally and mechanically. Their spontaneous interactions allow them to achieve the desired behavior. The main proposition assumes that the agent does not need a full model or to make inferences before taking actions; rather, the information necessary for any action can be derived from the environment with simple computations and very little internal state. In this paper, we present a simulation framework in which the agents are endowed with a sensing device, an oscillator network as controller and actuators to interact with the environment. The perception device is designed as an optic array emulating the principles of the animal retina, which assimilates stimuli resembling optic flow to be captured from the environment. The controller modulates informational variables to action variables in a sensory‐motor flow. Our approach is based on the Kuramoto model that describes mathematically a network of coupled phase oscillators and the use of evolutionary algorithms, which is proved to be capable of synthesizing minimal synchronization strategies based on the dynamical coupling between agents and environment. We carry out a comparative analysis with classical implementations taking into account several criteria. It is concluded that we should consider replacing the metaphor of symbolic information processing by that of sensory‐motor coordination in problems of multi‐agent organizations. Copyright

FAMAP: A Framework for Developing m-Health Apps

The edge-cutting mobile technologies have allowed the expansion of m-health applications for both patients and doctors. However, the variety of technologies, platforms and general-purpose development frameworks make developers and researchers to spend a considerable amount of time in developing m-health apps from scratch. This papers presents an ongoing research project about the creation of a framework for assisting developers and researchers in creating m-health apps called FAMAP. This framework is presented for the first time in the current article. Among others, this framework contains components for respectively (1) collecting data, (2) visualizing data analytics, (3) automating the definition and management of questionnaires, (4) implementing agent-based decision support systems and (5) supporting multi-modal communication. To show the utility of the proposed framework, this article presents some well-known and in-progress m-health apps developed with this framework. This work is assessed by considering (a) the usage data to show the commitment of users in one of the apps, and (b) the downloads and ranking in stores of another of the apps.

Exploring Criticality as a Generic Adaptive Mechanism

The activity of many biological and cognitive systems is not poised deep within a specific regime of activity. Instead, they operate near points of critical behavior located at the boundary between different phases. Certain authors link some of the properties of criticality with the ability of living systems to generate autonomous or intrinsically generated behavior. However, these claims remain highly speculative. In this paper, we intend to explore the connection between criticality and autonomous behavior through conceptual models that show how embodied agents may adapt themselves toward critical points. We propose to exploit maximum entropy models and their formal descriptions of indicators of criticality to present a learning model that drives generic agents toward critical points. Specifically, we derive such a learning model in an embodied Boltzmann machine by implementing a gradient ascent rule that maximizes the heat capacity of the controller in order to make the network maximally sensitive to external perturbations. We test and corroborate the model by implementing an embodied agent in the Mountain Car benchmark test, which is controlled by a Boltzmann machine that adjusts its weights according to the model. We find that the neural controller reaches an apparent point of criticality, which coincides with a transition point of the behavior of the agent between two regimes of behavior, maximizing the synergistic information between its sensors and the combination of hidden and motor neurons. Finally, we discuss the potential of our learning model to answer questions about the connection between criticality and the capabilities of living systems to autonomously generate intrinsic constraints on their behavior. We suggest that these “critical agents” are able to acquire flexible behavioral patterns that are useful for the development of successful strategies in different contexts.

A parsimonious model for locomotor in virtual agents based on dynamical coupling with the environment

In the domain of virtual agents modeling, computationalist approaches tend to predominate. The lack of plausibility of the rules in nature provides models with a low capacity to generalize and a very limited insight into behavioral phenomena. An alternative view is that behavior could emerge as a whole from the basic principles of non‐linear dynamics that underlie such behavior in the environment. This paper presents a dynamic agent endowed with a sensing device, a controller and actuators to interact with the environment. The control architecture is based on ordinary differential equations with the function of modulating the stimulus signals to action signals under a sensory‐motor flow. The parameter values are defined in an evolutionary process depending on the task to be performed. A series of experiments are presented to illustrate certain qualities of our model such as the adaptability to change, a highly intuitive and flexible design methodology, and a high degree of individual autonomy, among others. We think this kind of modeling is useful in the animation world for modeling flocks, crowds, and swarms as a valid alternative to other less parsimonious techniques.In the domain of virtual agents modeling, computationalist approaches tend to predominate. The lack of plausibility of the rules in nature provides models with a low capacity to generalize and a very limited insight into behavioral phenomena. An alternative view is that behavior could emerge as a whole from the basic principles of non-linear dynamics that underlie such behavior in the environment. This paper presents a dynamic agent endowed with a sensing device, a controller and actuators to interact with the environment. The control architecture is based on ordinary differential equations with the function of modulating the stimulus signals to action signals under a sensory-motor flow. The parameter values are defined in an evolutionary process depending on the task to be performed. A series of experiments are presented to illustrate certain qualities of our model such as the adaptability to change, a highly intuitive and flexible design methodology, and a high degree of individual autonomy, among others. We think this kind of modeling is useful in the animation world for modeling flocks, crowds, and swarms as a valid alternative to other less parsimonious techniques.

Extended neural metastability in an embodied model of sensorimotor coupling

The hypothesis that brain organization is based on mechanisms of metastable synchronization in neural assemblies has been popularized during the last decades of neuroscientific research. Nevertheless, the role of body and environment for understanding the functioning of metastable assemblies is frequently dismissed. The main goal of this paper is to investigate the contribution of sensorimotor coupling to neural and behavioral metastability using a minimal computational model of plastic neural ensembles embedded in a robotic agent in a behavioral preference task. Our hypothesis is that, under some conditions, the metastability of the system is not restricted to the brain but extends to the system composed by the interaction of brain, body and environment. We test this idea, comparing an agent in continuous interaction with its environment in a task demanding behavioral flexibility with an equivalent model from the point of view of “internalist neuroscience.” A statistical characterization of our model and tools from information theory allow us to show how (1) the bidirectional coupling between agent and environment brings the system closer to a regime of criticality and triggers the emergence of additional metastable states which are not found in the brain in isolation but extended to the whole system of sensorimotor interaction, (2) the synaptic plasticity of the agent is fundamental to sustain open structures in the neural controller of the agent flexibly engaging and disengaging different behavioral patterns that sustain sensorimotor metastable states, and (3) these extended metastable states emerge when the agent generates an asymmetrical circular loop of causal interaction with its environment, in which the agent responds to variability of the environment at fast timescales while acting over the environment at slow timescales, suggesting the constitution of the agent as an autonomous entity actively modulating its sensorimotor coupling with the world. We conclude with a reflection about how our results contribute in a more general way to current progress in neuroscientific research.