The purpose of this research is to contribute to social communication between humans and robots in scenes that have been considered difficult due to the limited facial expression capabilities of robots. In order to provide more detailed facial expressions, we designed a novel wire-driven 3D eyebrow using a soft material with a bending structure. We then demonstrated the mechanical properties of the eyebrow design and developed a prototype that could be implemented on the humanoid robot iCub to verify its operation. Lastly, we confirmed that the new design enables the production of more slight changes in facial expressions by allowing the eyebrows to change their shape at more delicate angles and continuity than the LED eyebrows of the conventional iCub.

Recognizing human partners is an essential social skill for building personalized and long-term human-robot interactions. However, robots deployed in complex, real-world environments have to face several challenges, such as managing unstructured interactions with multiple users, limited computational resources, and intrinsic and continuous variability of their sensory evidence. To cope with these challenges, we propose a framework to perform autonomous incremental learning for open-set face recognition suitable for unconstrained HRI scenarios. We validated the proposed framework in a real-world experiment, demonstrating its suitability to let the robot autonomously interact with multiple people while creating a labeled database of their faces across various encounters. Furthermore, we evaluated how an off-the-shelf model performed with data gathered from the HRI setting and proposed a fine-tuned model obtained with a transfer learning technique. Analyses about automatic threshold determination and rehearsal methods for memory sampling were also proposed. Our preliminary results suggest that exploiting the first-hand robot's experience could be crucial to ensure better models' performance and, therefore, could be advantageous for the acceptance and effectiveness of social robots in the long run. With this work, we aim to provide insights on continual learning approaches in the HRI field to promote autonomous and personalized solutions meaningful for real-world applications.

The ability to recognize human partners is an important social skill to build personalized and long-term Human-Robot Interactions (HRI). However, in HRI contexts, unfolding in ever-changing and realistic environments, the identification problem presents still significant challenges. Possible solutions consist of relying on a multimodal approach and making robots learn from their first-hand sensory data. To this aim, we propose a framework to allow robots to autonomously organize their sensory experience into a structured dataset suitable for person recognition during a multiparty interaction. Our results demonstrate the effectiveness of our approach and show that it is a promising solution in the quest of making robots more autonomous in their learning process.

Games are often used to foster human partners' engagement and natural behavior, even when they are played with or against robots. Therefore, beyond their entertainment value, games represent ideal interaction paradigms where to investigate natural human-robot interaction and to foster robots' diffusion in the society. However, most of the state-of-the-art games involving robots, are driven with a Wizard of Oz approach. To address this limitation, we present an end-to-end (E2E) architecture to enable the iCub robotic platform to autonomously lead an entertaining magic card trick with human partners. We demonstrate that with this architecture a robot is capable of autonomously directing the game from beginning to end. In particular, the robot could detect in real-time when the players lied in the description of one card in their hands (the secret card). In a validation experiment the robot achieved an accuracy of 88.2% (against a chance level of 16.6%) in detecting the secret card while the social interaction naturally unfolded. The results demonstrate the feasibility of our approach and its effectiveness in entertaining the players and maintaining their engagement. Additionally, we provide evidence on the possibility to detect important measures of the human partner`s inner state such as cognitive load related to lie creation with pupillometry in a short and ecological game-like interaction with a robot.

This paper proposes a method based on visual motion primitives to address the problem of action understanding. The approach builds in an unsupervised way a dictionary of kinematic primitives from a set of sub-movements obtained by segmenting the velocity profile of an action on the basis of local minima derived directly from the optical flow. The dictionary is then used to describe each sub-movement as a linear combination of atoms using sparse coding. The descriptive capability of the proposed motion representation is experimentally validated on the MoCA dataset, a collection of synchronized multi-view videos and motion capture data of cooking activities. The results show that the approach, despite its simplicity, has a good performance in action classification, especially when the motion primitives are combined over time. Also, the method is proved to be tolerant to view point changes, and can thus support cross-view action recognition. Overall, the method may be seen as a backbone of a general approach to action understanding, with potential applications in robotics.

Newborn infants are naturally attracted to human faces, a crucial source of information for social interaction. In robotics, acquisition of such information is crucial and social robots should also learn to exhibit such social skill. Deep learning algorithms are valid candidates to address the problem of face localisation. However, a major drawback of these methods is the large amount of data and human supervision needed in the training procedure. In this work, we propose a cognitive architecture to address autonomous learning from raw sensory signals without supervision. We demonstrate the success of our cognitive framework for the task of face localisation. The proposed cognitive architecture builds on existing work and uses audiovisual attention and a proactive stereo vision mechanism to autonomously direct a robot’s attentive focus towards human faces. The gathered information is used to incrementally generate a dataset that can be used to train a state-of-the-art deep network. The learning system imitates the typical learning process of infants and enhances the learning generalization process by leveraging on the interaction experience with people. The integration of HRI with machine learning, inspired by early development in humans, constitutes an innovative approach for improving autonomous learning in robots.

A social robot that is aware of our needs and continuously adapts its behaviour to them has the potential of creating a complex, personalized, human-like interaction of the kind we are used to have with our peers in our everyday lives. We are interested in exploring how would an adaptive architecture function and personalize to different users when given different initial values of its variables, i.e. when implementing the same adaptive framework with different robot personalities. Would an architecture that learns very quickly outperform a slower but steadier learning profile? To further explore this, we propose a cognitive architecture for the humanoid robot iCub supporting adaptability and we attempt to validate its functionality and test different robot profiles.

Authority and obedience are key regulatory elements in a society. Robots are becoming important part of our world, and are starting to interact in domains in which authority is an important aspect, as healthcare, teaching or law enforcement. Yet, there is little research on how people behave when robots show authority. In particular, although extensive investigations have been carried out on how authority can circumvent people's morality with experiments such as Milgram's or Stanford Prison, almost no research evaluated the effect of robots pushing the limits of people's own morality. This experiment tries to study this aspect by using a robot (geminoid) with the appearance, and thus authority of a famous person, and by pushing the boundaries asking morally controversial requests. The results show that, even though most people hesitate and recognize the requests as socially inappropriate, they obey to robots with authority. This suggests that the authority of the robot can push people to perform tasks usually considered as inappropriate.

editorial Open AccessHumane Robots—from Robots with a Humanoid Body to Robots with an Anthropomorphic Mind Share on Authors: Giulio Sandini Robotics, Brain and Cognitive Sciences, Istituto Italiano di Tecnologia, Genoa, Italy Robotics, Brain and Cognitive Sciences, Istituto Italiano di Tecnologia, Genoa, ItalyView Profile , Alessandra Sciutti Robotics, Brain and Cognitive Sciences, Istituto Italiano di Tecnologia, Genoa, Italy Robotics, Brain and Cognitive Sciences, Istituto Italiano di Tecnologia, Genoa, ItalyView Profile Authors Info & Affiliations ACM Transactions on Human-Robot InteractionVolume 7Issue 1May 2018 Article No.: 7pp 1–4https://doi.org/10.1145/3208954Published:16 May 2018 7citation726DownloadsMetricsTotal Citations7Total Downloads726Last 12 Months219Last 6 weeks31 Get Citation AlertsNew Citation Alert added!This alert has been successfully added and will be sent to:You will be notified whenever a record that you have chosen has been cited.To manage your alert preferences, click on the button below.Manage my AlertsNew Citation Alert!Please log in to your account Save to BinderSave to BinderCreate a New BinderNameCancelCreateExport CitationPublisher SiteeReaderPDF

Mutual synchronization plays a decisive role in effective collaborations in human joint tasks. Interaction between humans and robots need to show similar emergent coordination. To this aim models of human synchronization have recently been ported on collaborative robots with success [1]. However, it is also important to consider under which conditions the human partner is willing to adapt to the robot while performing a joint task. The main research goal of this study is to understand whether the temporal adaptation usually observed during human-human interaction occurs also during human-robot cooperation. We present a collaborative joint task engaging both human subjects and the humanoid robot iCub in pursuing an identical common goal: putting blocks into a box. We examine human action timing, evinced from motion capture data, in order to investigate whether humans adapt their behavior to the robot. We compare a quantitative measure of such adaptation with the subjective evaluation extracted from questionnaires. We observe that on average participants tend to adapt to their robotic partner. Nevertheless, by looking at individual behaviors, only few showed a clear adaptation to its timing, despite the vast majority of the subjects reported to have been influenced by the robot. We conclude discussing the potential factors influencing human adaptability, with the suggestion that the speed of execution of the robot is determinant in the coordination.

The ability to detect potentially interacting agents in the surrounding environment is acknowledged to be one of the first perceptual tasks developed by humans, supported by the ability to recognise biological motion. The precocity of this ability suggests that it might be based on rather simple motion properties, and it can be interpreted as an atomic building block of more complex perception tasks typical of interacting scenarios, as the understanding of non-verbal communication cues based on motion or the anticipation of others’ action goals. In this article, we propose a novel perspective for video analysis, bridging cognitive science and machine vision, which leverages the use of computational models of the perceptual primitives that are at the basis of biological motion perception in humans. Our work offers different contributions. In a first part, we propose an empirical formulation for the Two-Thirds Power Law , a well-known invariant law of human movement, and thoroughly discuss its readability in experimental settings of increasing complexity. In particular, we consider unconstrained video analysis scenarios, where, to the best of our knowledge, the invariant law has not found application so far. The achievements of this analysis pave the way for the second part of the work, in which we propose and evaluate a general representation scheme for biological motion characterisation to discriminate biological movements with respect to non-biological dynamic events in video sequences. The method is proposed as the first layer of a more complex architecture for behaviour analysis and human-machine interaction, providing in particular a new way to approach the problem of human action understanding.

Humans use eye gaze in their daily interaction with other humans. Humanoid robots, on the other hand, have not yet taken full advantage of this form of implicit communication. We designed a passive monocular gaze tracking system implemented on the iCub humanoid robot [Metta et al. 2008]. The validation of the system proved that it is a viable low-cost, calibration-free gaze tracking solution for humanoid platforms, with a mean absolute error of about 5 degrees on horizontal angle estimates. We also demonstrated the applicability of our system to human-robot collaborative tasks, showing that the eye gaze reading ability can enable successful implicit communication between humans and the robot.

Robots are at the position to become our everyday companions in the near future. Still, many hurdles need to be cleared to achieve this goal. One of them is the fact that robots are still not able to perceive some important communication cues naturally used by humans, e.g. gaze. In the recent past, eye gaze in robot perception was substituted by its proxy, head orientation. Such an approach is still adopted in many applications today. In this paper we introduce performance improvements to an eye tracking system we previously developed and use it to explore if this approximation is appropriate. More precisely, we compare the impact of the use of eye- or head-based gaze estimation in a human robot interaction experiment with the iCub robot and naïve subjects. We find that the possibility to exploit the richer information carried by eye gaze has a significant impact on the interaction. As a result, our eye tracking system allows for a more efficient human-robot collaboration than a comparable head tracking approach, according to both quantitative measures and subjective evaluation by the human participants.

It is generally accepted that a robot should exhibit a contingent behavior, adaptable to the needs of each individual user, to achieve a more natural and pleasant interaction. In this paper we have evaluated whether this general rule applies also when the robot plays a leading role and needs to motivate the human partner to keep a certain pace, as during training or teaching. Also among humans, in schools or factories, structured interaction is often guided by a predefined rhythm, which facilitates the coordination of the partners involved and is thought to maximize their efficiency. On the other hand, a pre-established timing forces all participants to adjust their natural speed to the external, sometimes not appropriate, timing requirement. Where does the optimal trade-off between these two paradigms lie? We have addressed this question in a dictation scenario where the humanoid robot iCub plays the role of a teacher and dictates brief English or Italian sentences to the participants. In particular we compare a condition in which the dictation is performed at a fixed timing with a condition in which iCub monitors subjects' gaze to adjust its dictation speed. The results are discussed both in terms of participants' subjective evaluation and their objective performance, by highlighting the advantages and drawbacks of the choice of contingent robot behavior.

Passing an object between two humans is a very natural and seamless operation, mainly thanks to non-verbal cues which facilitate the process. Just from action observation, humans can easily anticipate where and when a passing movement will end and how heavy the transported object is. But how could this natural understanding be ported to non-human agents? We introduce a simple robotic architecture to enable the iCub humanoid robot to visually recognize the weight of an object and select a lift-to-pass motion which implicitly communicates such information to the action partner. In this work we mainly focus on the building and training of the procedural memory module needed to store the association between the mass of an object and its visual appearance, and we propose how such a model can be used to successively select communicative lifting motions.

Knowing where a person is looking is an important parameter of every human-human interaction. Detecting a person's gaze could significantly improve the interaction capabilities of today's robotic agents. But many robots' visual systems are limited by data bandwidth and optical hardware. We propose a low-cost high-def pan/tilt/zoom active vision system that could significantly improve the robot's eye tracking capabilities. We tested the proposed system for improving mutual gaze detection in a human-robot interaction scenario and found significant results compared to systems without zoom capability.

Human interaction is based, among other factors, on non verbal and implicit communication. By observing the action of someone else we can automatically infer several non obvious details of what's happening, as the goal of the agent, his mood and even some features of the object he is using, e.g., if it is heavy or light. This action reading skill is developed very early in life and constitutes a fundamental basis for the development of collaboration. A similar capacity to implicitly communicate weight would be desirable also in a humanoid robot, to allow for a natural preparation for hand-over between the robot and the human partner. Here we have investigated how the ability to infer weight from human action observation develops during childhood and whether such ability generalizes to the observation of a humanoid robot. In particular, the robot was not programmed to perform human-like lifting actions, but just to replicate a property of human lifting deemed as determinant for weight reading: exhibiting a velocity proportional to object weight. Our results suggest that although 6-year-olds can already judge weight from the observation of a lifting action, they cannot generalize this skill to simplified robotic actions as the ones proposed here.

This paper presents a new technique to control highly redundant mechanical systems, such as humanoid robots. We take inspiration from two approaches. Prioritized control is a widespread multi-task technique in robotics and animation: tasks have strict priorities and they are satisfied only as long as they do not conflict with any higher-priority task. Optimal control instead formulates an optimization problem whose solution is either a feedback control policy or a feedforward trajectory of control inputs. We introduce strict priorities in multi-task optimal control problems, as an alternative to weighting task errors proportionally to their importance. This ensures the respect of the specified priorities, while avoiding numerical conditioning issues. We compared our approach with both prioritized control and optimal control with tests on a simulated robot with 11 degrees of freedom.

In this paper we tackle the problem of estimating the local compliance of tactile arrays exploiting global measurements from a single force and torque sensor. The proposed procedure exploits a transformation matrix (describing the relative position between the local tactile elements and the global force/torque measurements) to define a linear regression problem on the unknown local stiffness. Experiments have been conducted on the foot of the iCub robot, sensorized with a single force/torque sensor and a tactile array of 250 tactile elements (taxels) on the foot sole. Results show that a simple calibration procedure can be employed to estimate the stiffness parameters of virtual springs over a tactile array and to use these model to predict normal forces exerted on the array based only on the tactile feedback. Leveraging on previous works [1] the proposed procedure does not necessarily need a-priori information on the transformation matrix of the taxels which can be directly estimated from available measurements.

Seeing the world through the eyes of a child is always difficult. Designing a robot that might be liked and accepted by young users is therefore particularly complicated. We have investigated children's opinions on which features are most important in an interactive robot during a popular scientific event where we exhibited the iCub humanoid robot to a mixed public of various ages. From the observation of the participants' reactions to various robot demonstrations and from a dedicated ranking game, we found that children's requirements for a robot companion change sensibly with age. Before 9 years of age children give more relevance to a human-like appearance, while older kids and adults pay more attention to robot action skills. Additionally, the possibility to see and interact with a robot has an impact on children's judgments, especially convincing the youngest to consider also perceptual and motor abilities in a robot, rather than just its shape. These results suggest that robot design needs to take into account the different prior beliefs that children and adults might have when they see a robot with a human-like shape.

Cumulatively developing robots offer a unique opportunity to reenact the constant interplay between neural mechanisms related to learning, memory, prospection, and abstraction from the perspective of an integrated system that acts, learns, remembers, reasons, and makes mistakes. Situated within such interplay lie some of the computationally elusive and fundamental aspects of cognitive behavior: the ability to recall and flexibly exploit diverse experiences of one's past in the context of the present to realize goals, simulate the future, and keep learning further. This article is an adventurous exploration in this direction using a simple engaging scenario of how the humanoid iCub learns to construct the tallest possible stack given an arbitrary set of objects to play with. The learning takes place cumulatively, with the robot interacting with different objects (some previously experienced, some novel) in an open-ended fashion. Since the solution itself depends on what objects are available in the "now," multiple episodes of past experiences have to be remembered and creatively integrated in the context of the present to be successful. Starting from zero, where the robot knows nothing, we explore the computational basis of organization episodic memory in a cumulatively learning humanoid and address (1) how relevant past experiences can be reconstructed based on the present context, (2) how multiple stored episodic memories compete to survive in the neural space and not be forgotten, (3) how remembered past experiences can be combined with explorative actions to learn something new, and (4) how multiple remembered experiences can be recombined to generate novel behaviors (without exploration). Through the resulting behaviors of the robot as it builds, breaks, learns, and remembers, we emphasize that mechanisms of episodic memory are fundamental design features necessary to enable the survival of autonomous robots in a real world where neither everything can be known nor can everything be experienced.

This paper describes a developmental framework for action-driven perception in anthropomorphic robots. The key idea of the framework is that action generation develops the agent's perception of its own body and actions. Action-driven development is critical for identifying changing body parts and understanding the effects of actions in unknown or nonstationary environments. We embedded minimal knowledge into the robot's cognitive system in the form of motor synergies and actions to allow motor exploration. The robot voluntarily generates actions and develops the ability to perceive its own body and the effect that it generates on the environment. The robot, in addition, can compose this kind of learned primitives to perform complex actions and characterize them in terms of their sensory effects. After learning, the robot can recognize manipulative human behaviors with cross-modal anticipation for recovery of unavailable sensory modality, and reproduce the recognized actions afterward. We evaluated the proposed framework in the experiments with a real robot. In the experiments, we achieved autonomous body identification, learning of fixation, reaching and grasping actions, and developmental recognition of human actions as well as their reproduction.

The paper describes a developmental framework of action-driven perception in anthropomorphic robots. The key idea of the framework is that action develops the agent's perception of the own body and its action. In this framework, a robot voluntarily generates movements, and then develops the ability to perceive its own body and the effects of action primitives. The robot, moreover, demonstrates manipulative actions composed of the learned primitives, and characterizes the actions based on their sensory effects. After learning, the robot can predictively recognize humans' manipulative actions with cross-modal recovery of unavailable sensory information and reproduce the recognized actions. We evaluated the proposed framework in experiments with a real robot. In the experiments, we achieved developmental recognition of human actions as well as their reproduction.

We all `inhabit' continuously changing unstructured worlds where neither everything can be known nor can everything be experienced. The lifelong interplay between neural mechanisms associated with learning, memory, prospection and abstraction play a fundamental role in enabling cognitive agents to effortlessly connect their `past' with the `available present' and `possible future', most often in the context of realization of one's internal goals (and taking into account neurobiological and physical constraints related to one's embodiment). Investigation into the computational and neural basis underlying such mechanisms have value both from an intrinsic viewpoint of better understanding our own selves and at the same time creating artifacts that can flexibly assist us in the environments we inhabit and create (domestic, industrial, several others). Simply put, beyond a certain point a software programmer cannot travel the journey of a cognitive robot, they must learn to do it themselves! Enabling them to do so offers us with a unique opportunity to reenact the gradual process of infant developmental learning and investigate deeper into the underlying interplay between multiple fundamental cognitive processes from the perspective of an “integrated system” (human or humanoid that perceives, acts, learns, forgets and reasons). The article reports some advanced experiments conducted on the baby humanoid iCub to this effect, mainly 1. Learning to push and begin to anticipate how objects move, use this knowledge and 2. Learning to build the tallest possible stack given a random set of objects to play with (learning going on cumulatively in an open ended fashion). These playful experiments are used to summarize the gradually evolving cognitive architecture of the young EU funded project DARWIN.

In this paper we describe the mathematical model of a tendon driven eye. We focus on describing its movements, posing a specific attention on cyclotorsion, that is the rotation around the eye optical axis. This study aims at understanding the cause of a cyclotorsion effect that has been noticed on a real setup. The paper starts from a qualitative analysis of the effect. Then, it proposes two different models for its motion. Finally both models are validated by comparing their predictions with the outcomes on the real robot. Given the complexity of the system and of its motion results are for the moment just qualitative and quantitative comparison will be the goal of our future works.

In this paper we discuss the use of the infinite Gaussian mixture model and Dirichlet processes for learning robot movements from demonstrations. Starting point of this work is an earlier paper where the authors learn a non-linear dynamic robot movement model from a small number of observations. The model in that work is learned using a classical finite Gaussian mixture model (FGMM) where the Gaussian mixtures are appropriately constrained. The problem with this approach is that one needs to make a good guess for how many mixtures the FGMM should use. In this work, we generalize this approach to use an infinite Gaussian mixture model (IGMM) which does not have this limitation. Instead, the IGMM automatically finds the number of mixtures that are necessary to reflect the data complexity. For use in the context of a non-linear dynamic model, we develop a Constrained IGMM (CIGMM). We validate our algorithm on the same data that was used in [5], where the authors use motion capture devices to record the demonstrations. As further validation we test our approach on novel data acquired on our iCub in a different demonstration scenario in which the robot is physically driven by the human demonstrator.

Human arm muscular stiffness measurement is often a complex procedure which is of great interest for many disciplines from biomechanics to medicine and robotics. Modulation of impedance represents the principal mechanism underlying control of movements and interaction with external environment. Past literature proposed several methods to estimate multijoint hand stiffness while postural maintaining and dynamic tasks, mainly performed by means of planar robotic manipulanda. Despite these approaches are still considered robust and accurate, the computational burden of the robotic controller and hardware limitations make them not easy to implement. In the present paper a novel mechanism conceived for measuring multijoint planar stiffness by in single trial and in a reduced execution time is described and tested in different configurations. The device consisted in a mechanical rotary mechanism which applies cyclic radial perturbation to human arm of a known displacement and the force is acquired by means of a 6-axes commercial load cell. The outcomes suggest that the system is not only reliable in standalone mode but allows obtaining a reliable bi-dimensional estimation of arm stiffness even plugged in a planar manipulandum, dramatically reducing the amount of time for measurement and allowing to decouple the two controllers of the planar manipulator on which is mounted and the device itself.

We describe an interactive learning strategy that enables a humanoid robot to build a representation of its workspace: we call it a Reachable Space Map. The robot learns this map autonomously and online during the execution of goal-directed reaching movements; reaching control is based on kinematic models that are learned online as well. The map can be used to estimate the reachability of a fixated object and to plan preparatory movements (e.g. bending or rotating the waist) that improve the effectiveness of the subsequent reaching action. Three main concepts make our solution innovative with respect to previous works: the use of a gaze-centered motor representation to describe the robot workspace, the primary role of action in building and representing knowledge (i.e. interactive learning), the realization of autonomous online learning. We evaluate our strategy by learning the workspace of a simulated humanoid robot and we show how this knowledge can be exploited to plan and execute complex actions, like whole-body bimanual reaching.

The paper describes a constructive approach on active perception for anthropomorphic robots. The key idea is that a robot tries to identify a human's action as an own action based on the observation of action effects for objects. In the proposed framework, the active perception is decomposed into the three phases; First, a robot voluntarily generates actions to discover the own body and objects. Second, the robot characterizes its own action with the effect for the objects. Third, the robot identifies the human action with the own action. The mirrored perception of the own action and the human's action allows the robot to share the goal-directed behavior with humans. The proposed framework of active perception was experimentally validated with the integrated sensory modalities of vision, proprioception and touch.

Muscle co-contraction can be modeled as an active modulation of the passive musculo-skeletal compliance. Within this context, recent findings in human motor control have shown that active compliance modulation is fundamental when planning movements in presence of unpredictability and uncertainties. Along this line of research, this paper investigates the link between active impedance control and unpredictability, with special focus on robotic applications. Different types of actuators are considered and confronted to extreme situations such as moving in an unstable force field and controlling a system with significant delays in the feedback loop. We use tools from stochastic optimal control to illustrate the possibility of optimally planning the intrinsic system stiffness when performing movements in such situations. In the extreme case of total feedback absence, different actuators model are considered and their performance in dealing with unpredictability compared. Finally, an application of the proposed theories on planning reaching movements with the iCub humanoid platform is proposed.

When a robot is required to safely interact with a physical environment, two approaches are typically reported in literature: using a force/torque sensor to regulate the interaction forces at the end effector, or integrating sensors in each robot joint to regulate their torques. In this paper we want to discuss the benefits and the disadvantages of the two approaches, showing a direct comparison between the information which can be obtained from the two categories of sensors. Results obtained on the new iCub arm, which integrates torque sensing capabilities at joint level will be presented and discussed.

This paper describes the control of a human-like robotic neck actuated with tendons. The controller regulates the length of the tendons to achieve a desired orientation of the neck and at the same time it maintains the tension of the tendons within certain limits. The solution we propose does not use any model of the system, but it relies on online learning of the different Jacobian mappings required by the controller. Learning, data acquisition and control are simultaneous; thus learning is completely autonomous, and purely online. We show that after enough iterations the controller produces straight trajectories in the task space and is able to maintain the tension of the tendons within safe limits.

During the last decades, interaction (with humans and with the environment) has become an increasingly interesting topic of research within the field of robotics. At the basis of interaction, a fundamental role is played by the ability to actively regulate the interaction forces. In this paper we propose a technique for controlling the interaction forces exploiting a proximal six axes force/torque sensor. The major assumption is the knowledge of the point where external forces are applied. The proposed approach is tested and validated on the four limbs of the iCub, a humanoid robot designed for research in embodied cognition. Remarkably, the proposed approach can be used to implement active compliance in other non passively back-drivable manipulators by simply inserting one or more force/torque sensor anywhere along the kinematic chain.

Online optimal planning of robotic arm movement is addressed. Optimality is inspired by computational models, where a “cost function” is used to describe limb motions according to different criteria. A method is proposed to implement optimal planning in Cartesian space, minimizing some cost function, by means of numerical approximation to a generalized nonlinear model predictive control problem. The Extended RItz Method is applied as a functional approximation technique. Differently from other approaches, the proposed technique can be applied on platforms with strict control temporal constraints and limited processing capability, since the computational burden is completely concentrated in an off-line phase. The trajectory generation on-line is therefore computationally efficient. Task to joint space conversion is implemented on-line by a closed loop inverse kinematics algorithm, taking into account the robot's physical limits. Experimental results, where a 4DOF arm moves according to a particular nonlinear cost, show the effectiveness of the proposed approach, and suggest interesting future developments.

In this paper we describe the design of a Cartesian Controller for a generic robot manipulator. We address some of the challenges that are typically encountered in the field of humanoid robotics. The solution we propose deals with a large number of degrees of freedom, produce smooth, human-like motion and is able to compute the trajectory on-line. In this paper we support the idea that to produce significant advancements in the field of robotics it is important to compare different approaches not only at the theoretical level but also at the implementation level. For this reason we test our software on the iCub platform and compare its performance against other available solutions.

This work proposes a plausible approach for a humanoid robot to define its own body parts based on the correlation of two different sensory signals: vision and proprioception. The high correlation between the motions in vision and proprioception informs the robot that the visually attractive object is related to the motor function of its own body. When the robot finds the highly motor-correlated object during head-arm movements, the visuomotor cues such as the body posture and visual features are stored in a visuomotor memory. Then, developmentally, the robot defines the motor-correlated objects as the own-body parts without prior knowledge on the body appearances and kinematics. It is also adaptable to extended body parts such as a grasped tool. The body movements are generated by stochastic motor babbling. The visuomotor memory biases the babbling to keep the own-body parts in sight. This memory-based bias towards the own-body parts helps the robot explore the large head-arm joint space. The acquired visuomotor memory is also used to anticipate the own-body image from the motor commands in advance of the body movement. The proposed approach was evaluated with two humanoid platforms; iCub and James.

Starting from human ¿sense of touch,¿ this paper reviews the state of tactile sensing in robotics. The physiology, coding, and transferring tactile data and perceptual importance of the ¿sense of touch¿ in humans are discussed. Following this, a number of design hints derived for robotic tactile sensing are presented. Various technologies and transduction methods used to improve the touch sense capability of robots are presented. Tactile sensing, focused to fingertips and hands until past decade or so, has now been extended to whole body, even though many issues remain open. Trend and methods to develop tactile sensing arrays for various body sites are presented. Finally, various system issues that keep tactile sensing away from widespread utility are discussed.

This paper presents the design, and performance of a high fidelity three degree-of-freedom wrist exoskeleton robot, for neuroscience study, training and rehabilitation. The IIT-Wrist is intended to provide kinesthetic feedback during the training of motor skills or rehabilitation of reaching movements. Motivation for such applications is based on findings that show robot-assisted physical therapy aids in the rehabilitation process following neurological injuriesIn the present paper the IIT-Wrist haptic robot is described in terms of kinematics and haptics features to meet specific requirements for a safety human.machine interaction. In relation with a feasibility study in the field of robot therapy a preliminary training of stroke patient was perfrormed. The task consisted in tracking a target using one degree of freedom at time: Flexion/Extension, adduction/abduction, pronation/supination separately. The target motion is harmonic and tracking is aided by a suitable force field. The preliminary study with three patients shows the stability and the efficacy of the control scheme.

The state-of-the-art in control of hand prosthetics is far from optimal. The main control interface is represented by surface electromyography (EMG): the activation potentials of the remnants of large muscles of the stump are used in a non-natural way to control one or, at best, two degrees-of-freedom. This has two drawbacks: first, the dexterity of the prosthesis is limited, leading to poor interaction with the environment; second, the patient undergoes a long training time. As more dexterous hand prostheses are put on the market, the need for a finer and more natural control arises. Machine learning can be employed to this end. A desired feature is that of providing a pre-trained model to the patient, so that a quicker and better interaction can be obtained. To this end we propose model adaptation with least-squares SVMs, a technique that allows the automatic tuning of the degree of adaptation. We test the effectiveness of the approach on a database of EMG signals gathered from human subjects. We show that, when pre-trained models are used, the number of training samples needed to reach a certain performance is reduced, and the overall performance is increased, compared to what would be achieved by starting from scratch.

For a complex autonomous robotic system such as a humanoid robot, motor-babbling-based sensorimotor learning is considered an effective method to develop an internal model of the self-body and the environment autonomously. However, learning process requires much time for exploration and computation. In this paper, we propose a method of sensorimotor learning which explores the learning domain actively. Our approach discovers that the embodied learning system can design its own learning process actively, which is different from the conventional passive data-access machine learning. The proposed model is characterized by a function we call the ¿confidence¿, and is a measure of the reliability of state control. The confidence for the state can be a good measure to bias the exploration strategy of data sampling, and to direct its attention to areas of learning interest. We consider the confidence function to be a first step toward an active behavior design for autonomous environment adaptation. The approach was experimentally validated in typical sensorimotor coordination such as arm reaching and object fixation, using the humanoid robot James and the iCub simulator.

The dexterity of active hand prosthetics is limited not only due to the limited availability of dexterous prosthetic hands, but mainly due to limitations in interfaces. How is an amputee supposed to command the prosthesis what to do (i.e., how to grasp an object) and with what force (i.e., holding a hammer or grasping an egg)? So far, in literature, the most interesting results have been achieved by applying machine learning to forearm surface electromyography (EMG) to classify finger movements; but this approach lacks, in general, the possibility of quantitatively determining the force applied during the grasping act. In this paper we address the issue by applying machine learning to the problem of regression from the EMG signal to the force a human subject is applying to a force sensor. A detailed comparative analysis among three different machine learning approaches (Neural Networks, Support Vector Machines and Locally Weighted Projection Regression) reveals that the type of grasp can be reconstructed with an average accuracy of 90%, and the applied force can be predicted with an average error of 10%, corresponding to about 5N over a range of 50N. None of the tested approaches clearly outperforms the others, which seems to indicate that machine learning as a whole is a viable approach.

In this paper, we describe the implementation of a precise reaching controller on an upper-torso humanoid robot. The solution we propose does not rely on prior models of the kinematic structure of either the arm or the head. A learning strategy enables the robot to acquire the required sensory-motor transformations. After learning the robot is able to precisely reach for a visually identified object in the 3-dimensional space. In this technique we use the fixation point (represented in the head joints motor space) as a reference frame to code the position of the object and to represent the eye-to-hand Jacobian matrix. This strategy successfully deals with the kinematic redundancy of the structure and constraints the dimensionality of the problem.

This survey presents an overview of the autonomous development of mental capabilities in computational agents. It does so based on a characterization of cognitive systems as systems which exhibit adaptive, anticipatory, and purposive goal-directed behavior. We present a broad survey of the various paradigms of cognition, addressing cognitivist (physical symbol systems) approaches, emergent systems approaches, encompassing connectionist, dynamical, and enactive systems, and also efforts to combine the two in hybrid systems. We then review several cognitive architectures drawn from these paradigms. In each of these areas, we highlight the implications and attendant problems of adopting a developmental approach, both from phylogenetic and ontogenetic points of view. We conclude with a summary of the key architectural features that systems capable of autonomous development of mental capabilities should exhibit

This work discusses the design and control approach of a 3 DOF waist joint for the "iCub" humanoid robot. "iCub" is a child like crawling robot that resembles a 2 and a half year old child. The ultimate goal of this project is to provide the cognition research community with an open human like platform for understanding of cognitive systems through the study of cognitive development. The designs of the mechanisms adopted for the waist joint are discussed. This is accompanied by discussion of the control scheme design and presentation of experimental results showing the performance of the mechanism

Humans perceive the world using five senses. Research results suggest that this multisensorial perception may be of fundamental importance for development and learning, as well as for creating cognitive capabilities. Moreover, humans have the capacity to create intersensorial expectations to guide attention and perception. We are interested in comprehending how these capabilities may improve robot perception. In this line of research, we present a cross-modal perceptual architecture that can segment objects based on visual-auditory sensorial cues, construct an associative sound-object memory, and create visual expectations of objects (attentional priming) using a sound recognition algorithm.

Within the field of Neuro Robotics we are driven primarily by the desire to understand how humans and animals live and grow and solve every day's problems. To this aim we adopted a "learn by doing" approach by building artificial systems, e.g. robots that not only look like human beings but also represent a model of some brain process. They should, ideally, behave and interact like human beings (being situated). The main emphasis in robotics has been on systems that act as a reaction to an external stimulus (e.g. tracking, reaching), rather than as a result of an internal drive to explore or "understand" the environment. We think it is now appropriate to try to move from acting, in the sense explained above, to "understanding". As a starting point we addressed the problem of learning about the effects and consequences of self-generated actions. How does the robot learn how to pull an object toward itself or to push it away? How does the robot learn that spherical objects roll while a cube only slides if pushed? Interacting with objects is important because it implicitly explores object representation, event understanding, and can provide definition of object-hood that could not be grasped with a mere passive observation of the world. Further, learning to understand what one's own body can do is an essential step toward learning by imitation. In this view two actions are similar not only if their kinematics and dynamics are similar but rather if the effects on the external world are the same. Along this line of research we discuss some recent experiments performed at the AI-Lab at MIT and at the LIRA-Lab at the University of Genova on COG and Babybot respectively. We show how the humanoid robots can learn how to poke and prod objects to obtain a consistently repeatable effect (e.g. sliding in a given direction), to help visual segmentation, and to interpret a poking action performed by a human manipulator.

Developmental robotics is an emerging field located at the intersection of robotics, cognitive science and developmental sciences. This paper elucidates the main reasons and key motivations behind the convergence of fields with seemingly disparate interests, and shows why developmental robotics might prove to be beneficial for all fields involved. The methodology advocated is synthetic and two-pronged: on the one hand, it employs robots to instantiate models originating from developmental sciences; on the other hand, it aims to develop better robotic systems by exploiting insights gained from studies on ontogenetic development. This paper gives a survey of the relevant research issues and points to some future research directions.

This paper describes a novel network model, which is able to control its growth on the basis of the approximation requests. Two classes of self-tuning neural models are considered; namely Growing Neural Gas (GNG) and SoftMax function networks. We combined the two models into a new one: hence the name GNG-Soft networks. The resulting model is characterized by the effectiveness of the GNG in distributing the units within the input space and the approximation properties of SoftMax functions. We devised a method to estimate the approximation error in an incremental fashion. This measure has been used to tune the network growth rate. Results showing the performance of the network in a real-world robotic experiment are reported.

In the classical neural-gas method, the neurons are located in the input space according to the input density. But if we want to use this kind of approach as a function approximator, it can be interesting to change the spatial distribution of the neurons, so that they concentrate in regions where the unknown function appeared to be more complex. To achieve this goal, we modified the update rule of the neurons so that they move towards regions where the estimated local error is high. In this article, we first show how to estimate this error locally to each neuron, and then we detail the modification of the algorithm. Finally, we present some simulations results that allow us to compare the modified approach with the classical one.

We propose a new neural approach for approximating function using a reinforcement-type learning: each time the network generates an output, the environment responds with the scalar distance between the delivered output and the expected one. Thus, this distance is the only information the network can use to modify the estimation of the multi-dimensional output. This reinforcement feature is embedded in a neural-gas method, taking advantages of the different facilities it offers. We detail the global algorithm and we present some simulation results in order to show the behaviour of the developed method.

Visual servoing based upon geometrical features such as image points coordinates is now well set on. Nevertheless, this approach has the drawback that it usually needs visual marks on the observed object to retrieve geometric features. The idea developed here is that these features can be retrieved by integrating dynamic ones, which can be estimated without any a priori knowledge of the scene. Thus, more realistic scenes can be used to achieve vision-based control such as positioning tasks. We show that an affine model is insufficient to ensure convergence and that a quadratic one is needed. Finally, results are presented related to the positioning with respect to a complex scene.

The aim of the paper is to propose that the adoption of a framework of biological development is suitable for the construction of artificial systems. We argue that a developmental approach does provide unique insights on how to build highly complex and adaptable artificial systems. In turn, it might also aid neuroscientists in a better understanding of the human brain functions involved in sensori-motor control. To illustrate our point, we use as an example the acquisition of goal-directed reaching in human infants, and demonstrate the underlying mechanisms of biological development. In a second part, we outline a) how mechanisms of biological development can be adapted to the artificial world, and b) how this artificial development differs from traditional engineering approaches to robotics. The experimental part is based on a set-up composed of a monocular robot head with two degrees of freedom and a two degrees of freedom arm. The motor control of both the head and the arm is based on the so-called force field approach described in the biological literature as a mechanism implemented at the level of the spinal cord to control ballistic motion of the arm. The visual part is, at this stage, limited to a simple target tracking procedure based on color information. The goal of the system is to reach with the arm the target fixated by the eye. Initially the new-born system is capable of performing a limited number of motor actions coded as rough visuo-motor maps and initiated by the appearance of the target in the field of view. During the developmental phase the system, through repetitive reaching of different points in the arm workspace, refines the visuo-motor maps without explicit knowledge of the system's kinematic parameters.

Vergence provides robot vision systems with a crucial degree of freedom: it enables fixation of points in visual space at different distances from the observer. Moreover, the control of vergence definitely influences the degree of binocular fusion in an active vision system. The paper shows how binocular fusion can be considered as a basic visual cue to elicit vergence and smooth pursuit movements. The choice of this visual cue together with the log-polar image plane geometry turns out to be very relevant in the design of a binocular tracking system. Real-time experiments prove the effectiveness of the proposed approach.

Vergence control enables fixation of points in visual space at different distances from the observer. It must provide not only a stable fixation, but a stable binocular fusion and a fast and accurate reaction to changes in the environment. To obtain this kind of performance the paper focuses on the use of two dynamic quantities as inputs for the vergence control process: the temporal evolution of the degree of fusion and the rate of expansion/contraction of image patterns as a consequence of the relative motion between an object and the observer. The use of space variant image geometry is proposed and some advantages of this approach are discussed in order to demonstrate the relevance of log-polar image geometry in relation to vergence control. Three experiments are described: the first concentrates attention on the results of visual processing algorithms in an open loop situation; the other two, illustrate the behavior of the real-time vergence system in response to continuous changes of depth of a controlled target.

The first application of a new generation of space variant sampling visual sensor is described in this paper. Time to contact computation from optic flow is performed using a very simple system consisting of a space variant chip connected through a parallel port to a 486 based PC. Results of experiments will be reported.

This paper describes a novel approach to the problem of reaching an object in space under visual guidance. The approach is characterized by a great robustness to calibration errors, such that virtually no calibration is required. Servoing is based on binocular vision: a continuous measure of the end-effector motion field, derived from real-time computation of the binocular optical flow over the stereo images, is compared with the actual position of the target and the relative error in the end-effector trajectory is continuously corrected. The paper outlines the general framework of the approach, shows how visual measures are obtained and discusses the synthesis of the controller along with its stability analysis. Real-time experiments are presented to show the applicability of the approach in real 3-D applications.

The aim of this paper is to discuss and present a navigation architecture for an autonomous mobile robot, based on the integration of ultrasonic sensors and stereo vision. The distinctive features are the integration of a "classical" planning approach and subsumption architecture for obstacle avoidance, and the fusion of visual and ultrasonic sensory data. The goal of the experiment is to plan missions between two user defined points in the environment, autonomously avoiding static and dynamic obstacles along the planned path.>

The role of vision in two-arm manipulation is investigated in order to suggest a qualitative approach for the solution of coordinated manipulative actions. The visuomotor behavior proposed, although partially supported by experimental results, differs from more geometrical approaches based on explicit 4D measures and trajectory estimation in that it is based on direct visual information, such as stereo disparity or the 2D image motion field. The experiment is limited to analysis of the optical flow information extracted from manipulative actions performed by a human operator. These experiments demonstrate that the qualitative approach proposed is not in contrast with human behavior and that the two-arm manipulation task proposed can be seen as a testbed for qualitative and geometrical approaches.>

A qualitative approach to visually guided navigation based on the computation of optical flow field is presented. The approach is based on the use of two cameras mounted on a mobile robot and with the optical axis directed in opposite directions, such that the two visual fields do not overlap (divergent stereo). Range computation is based on the computation of the apparent image speed on images acquired during the robot's motion. An example of reflex-type control of motion, driven by differential estimation of the flow field measured by the two eyes, is presented. It is shown how a difficult task like navigating through a funneled corridor with obstacles is possible without the need for metric depth estimation.>

The aim of this paper is to discuss and present some experimental data on the use of optical, chemical and ultrasound sensors to control the motion of single and cooperating robots. The distinctive feature is that in both cases the motion of the robots is not driven by a localized target (e.g. a light, or a sound) but by a concentration gradient such as the one generated by a gas leak. Experiments using a simple optical sensor reading a variable density pattern of dark spots painted on the ground plane, are presented along with experiments performed using simple chemical sensors measuring the local gradient of a substance. Different strategies are discussed based on different configurations of the sensing devices and reflexive motor controls. The goal of the experiment is to analyze the performance of simple autonomous robots in reaching the point of highest concentration (i.e. the gas leak). This performance has been analyzed for single robots and for cooperating robots using a very simple communication channel.

This work presents some experiments of a real-time navigation system driven by two cameras pointed laterally to the navigation direction (divergent stereo). The approach is based on the observation that the stereo set-up traditionally used in vision (i.e. with the optical axis pointing forward) may not be the best one for navigation, and particularly for continuous control of a mobile actor moving in unconstrained environment. For navigation purposes, the driving information is not distance (as it is obtainable by a stereo set-up) but motion and, more precisely, by optical flow information computed over different areas of the visual field. Following this idea, a mobile vehicle has been equipped with a pair of cameras looking laterally (much like honeybees) and a controller based on fast, real-time computation of optical flow, has been implemented. The control of the mobile robot (ROBEE) is based on the comparison between the apparent image speed of the left and the right eye. A detailed description of the control structure is presented to demonstrate the feasibility of the approach in driving the mobile robot within a very cluttered environment. A discussion about the potentialities of the approach and the implications in terms of sensor's structure is also presented.

The application of an anthropomorphic retina-like visual sensor and the advantages of polar and log-polar mapping for visual navigation are investigated. It is demonstrated that the motion equations that relate the egomotion and/or the motion of the objects in the scene to the optical flow are considerably simplified if the velocity is represented in a polar or log-polar coordinate system, as opposed to a Cartesian representation. The analysis is conducted for tracking egomotion but is then generalized to arbitrary sensor and object motion. The main result stems from the abundance of equations that can be written directly that relate the polar or log-polar optical flow with the time to impact. Experiments performed on images acquired from real scenes are presented.>

The paper describes the practical implementation of a vision-based navigation system for a mobile robot operating in indoor environments. The robot acquires visual information by means of three CCD cameras mounted on board. A stereo pair is used for ground plane obstacle detection and avoidance, while the third camera is used to locate landmarks and compute the robot's position. Odometric readings are used to guide visual perception by simple 'where to look next' strategies. The whole processing and the control architecture determining the overall behaviour of the robot are mainly implemented on a parallel MIMD machine. Some examples are presented, showing how the robot moves in a partially structured environment reaching the specified goal points with a fair degree of accuracy, avoiding unpredicted obstacles and following trajectories obtained through the cooperation of the various navigation modules running in parallel.>

A cooperative schema is proposed in which binocular disparity, computed on several stereo images over time, is combined with optical flow from the same sequence to obtain a relative-depth map of the scene. Both time-to-impact and depth scaled by the distance of the camera from the fixation point in space are considered as good, relative measurements which are based on the viewer (but centered on the environment). Two experiments, performed on image sequences from real scenes, are presented.>

A description is presented of the social organization of societies of cellular mobile units featuring instinctive behavior. Each robotic unit has its own personality and lives independently from the others. Useful tasks are carried out through collaboration rather than by individual effort. The behavior of each unit derives from a subsumption-like control structure, which emphasizes the roles of innate personality, external stimuli, and communication. A number of different robotic personalities are described and techniques of implementing them in real robot units are outlined. The implementation of instinctive behavior is described for the case of a robotic vehicle system (ROBBIE).>

One of the main problems in robotic research is planning. Different approaches have been considered, ranging from global to local planning. In the authors' approach, the key idea is that of planning movements of redundant robots, which involves both robotic and AI aspects. Pushing is used as a case study and sensorial feedback is considered in order to provide the planner with updated information on the dynamic evolution of the scene and to be able to deal with a priori unknown objects.>

Present a perspective in relation to the use of vision for the control of robots and for the monitoring of robot actions. In order to explain the approach some experiments have been performed in the field of object manipulation and robot navigation. In particular the monitoring of pushing and tapping actions by means of dynamic measurements performed on image sequences are reported. The experiments show that, not only it seems possible to detect unexpected events, but it is possible to control the arm or vehicle trajectory in such a way to achieve a complex task like the pushing of an unknown object along a predetermined trajectory.>

The use of an anthropomorphic, retinalike visual sensor for navigation tasks is investigated. The main advantage, besides the topological scaling and rotation invariance, stems from the considerable data reduction obtained with nonuniform sampling, in conjunction with high resolution in the part of the field of view corresponding to the focus of attention. Active movements are also considered to be a beneficial feature, solving the depth-from-motion problem and maintaining a 3-D representation of the viewed scene. For short range navigation, a tracking egomotion strategy is adopted which greatly simplifies the motion equations and complements the characteristics of the retinal sensor. An algorithm for the computation of depth from motion is developed for image sequences acquired with the retinal sensor, and an error analysis is carried out to determine the uncertainty of range measurements. An experiment is presented in which depth maps are computed from a sequence of images sampled with the retinalike sensor, building a volumetric representation of the scene.>

The extraction of depth information from a sequence of images is investigated. An algorithm that exploits the constraint imposed by active motion of the camera is described. Within this framework, in order to facilitate measurement of the navigation parameters, a constrained egomotion strategy was adopted in which the position of the fixation point is stabilized during the navigation (in an anthropomorphic fashion). This constraint reduces the dimensionality of the parameter space without increasing the complexity of the equations. A further distinctive point is the use of two sampling rates: the faster (related to the computation of the instantaneous optical flow) is fast enough to allow the local operator to sense the passing edge (or, in other words, to allow the tracking of moving contour points), while the slower (used to perform the triangulation procedure necessary to derive depth) is slow enough to provide a sufficiently large baseline for triangulation. Experimental results on real image sequences are presented.>

The extraction of reliable range data from images is investigated, considering, as a possible solution, the integration of different sensor modalities. Two different algorithms are used to obtain independent estimates of depth from a sequence of stereo images. The results are integrated on the basis of the uncertainty of each measure. The stereo algorithm uses a coarse-to-fine control strategy to compute disparity. An algorithm for depth-from-motion is used, exploiting the constraint imposed by active motion of the cameras. To obtain a 3D description of the objects, the motion of the cameras is purposefully controlled, in such a manner as to move around the objects in view while the gaze is directed toward a fixed point in space. This egomotion strategy, which is similar to that adopted by the human visuomotor system, allows a better exploration of partially occluded objects and simplifies the motion equations. When tested on real scenes, the algorithm demonstrated a low sensitivity to image noise, mainly due to the integration of independent measures. An experiment performed on a real scene containing several objects is presented.>

Abstract Image understanding is concerned with the elucidation of a computational base inherent in perceiving a three‐dimensional world using vision. This paper describes a low‐level (or early) vision software system, developed in the context of current collaborative research activities in image understanding, which goes some way toward fulfilling the goals of portability, ease of use, and general‐purpose extensibility. Since visual perception uses several types of disparate, but interrelated, information in some explicit cognitive organization, a central objective of the work is to represent this information in a coherent integrated manner which allows one interactively to investigate the properties of the interdependency between information types.

In this paper a contour based motion estimation technique is presented applied to the measurement of the optic-flow field from image sequences. In particular two experimental situations are considered: the former, in which a fixed camera analyzes a scene with some moving objects in order to detect their trajectories; the latter, in which a sensor moves in a static environment, determining the depth map of the "world". For the first case, we compute for every contour the prevalent direction of movement. For the second, we solve the problem of recovering the 3D structure of the worldby computing the depth field from a sequence of images acquired during a motion of the camera, controlled in order to mantain the fixation point (i.e. the point projected over the image plane) still during the entire sequence. The algorithm has proved to be very robust in real scene analysis and an acceptable time for computation is required. Some experimental results are presented.

Interaction between nerve cells, mediated by electrical coupling, is a common feature of the nervous system. When such electrical interactions occur within an entire population of similar neurons, the dynamic behavior of each individual cell is influenced by the properties of the whole network. It is shown that when the membrane impedance of such cells can be linearized to a satisfactory approximation and the zeros and poles of the linearized impedance are known, analytical solutions to the network equations can be found that describe the dynamics of the responses of any given cell to one-dimensional and two-dimensional stimuli. This analysis applied to the network of photoreceptors and horizontal cells in the vertebrate retina may be helpful in understanding the initial stages in the processing of visual information.

The aim of this paper is to outline a model of the eye-head system and some of its control strategies. To formulate such a model, the eye-head system was stimulated with random or periodic visual targets. Coupling between the eye and head motor commands was evident in the experiments performed, and this was also found to be the case for acoustic and tactile stimuli. Further experiments were performed to investigate whether the central nervous system (CNS) would be able to overcome the relative lack offlexibility shown by the model of the eye-head control system initially proposed. These experiments, in which it was required to execute independent eye-head movements towards two simultaneous targets (visual and acoustic), showed an impaired performance, indicating that an overcoming action by the CNS is not always possible. The findings are interpreted in terms of an interaction between a "hardware" low-level controller and a "software" high-level decisionmaker.