Psychologist - Psychotherapist - PhD in psychology. - Program Manager in Scientific Research and Knowledge dissemination @ META-LAB: Laboratory of MEthodologies and Technologies for Autism. Institute for Research, Education and information on disabilities Address: Naples
In this paper we describe a new approach in Evolutionary Robotics according to which human breede... more In this paper we describe a new approach in Evolutionary Robotics according to which human breeders are involved in the evo- lutionary process. While traditionally robots are selected to reproduce automatically according to a fltness formula, a quantitative and strictly deflned measure, human breeders can operate a selection based on qual- itative criteria, rewarding behaviors that can slip between the meshes woven by the fltness formula. In the authors' opinion this may bring advantages to the Evolutionary Robotics methodology, allowing for the production of robots that display more and more multiform behavior. In order to illustrate this approach, a piece of software, BreedBot, was developed in which human breeders can intervene in evolving robots, complementing the automatic evaluation. After describing the software, some results on sample evolutionary processes are reported showing that the joint use of human and artiflcial selection on an exploration task generates robots ...
Particolarmente originali e attraenti sono i giochi per bambini di tipo “playware”, che prevedono... more Particolarmente originali e attraenti sono i giochi per bambini di tipo “playware”, che prevedono l’uso di tecnologie pervasive volte a favorire lo sviluppo di nuovi scenari ludici dove il giocare affina contemporaneamente le abilità cognitive, motorie e sociali. Il gioco consiste in una corsa tra robot fuori dal comune poiché il robot vincitore viene decretato in base alle azioni che il giocatore compie nel mondo reale.
Robots are attractive for public as they represent a clear and ambitious scientific challenge: an... more Robots are attractive for public as they represent a clear and ambitious scientific challenge: an artificial being creation similar to natural ones by man. The popular enthusiasm grows up in parallel with huge scientific research progress in this field. On this success wave many Science Centers, Science Museums and Science Festivals propose spaces and moments to spread theoretic, methodological and technical issues of this discipline. In this paper we describe some of our exhibits prototypes that were designed and carried out to communicate main approaches of modern robotics (telerobotics, cognitive robotics, autonomous and evolutionary robotics and collective robotics). These exhibits have been conceived as laboratories where a visitor can experiment and put the hands on various kinds of robot. They have been accomplished by putting together cheap and easily available materials. This way it has been possible to present these exhibits in many important scientific divulgation events throughout Europe with a relatively modest expense.
When monkeys tackle novel complex behavioral tasks by trial-and-error they select actions from re... more When monkeys tackle novel complex behavioral tasks by trial-and-error they select actions from repertoires of sensorimotor primitives that allow them to search solutions in a space which is coarser than the space of fine movements. Neuroscientific findings suggested that upper-limb sensorimotor primitives might be encoded, in terms of the final goal-postures they pursue, in premotor cortex. A previous work by the authors reproduced these results in a model based on the idea that cortical pathways learn sensorimotor primitives while basal ganglia learn to assemble and trigger them to pursue complex reward-based goals. This paper extends that model in several directions: a) it uses a Kohonen network to create a neural map with population encoding of postural primitives; b) it proposes an actor-critic reinforcement learning algorithm capable of learning to select those primitives in a biologically plausible fashion (i.e., through a dynamic competition between postures); c) it proposes a procedure to pre-train the actor to select promising primitives when tackling novel reinforcement learning tasks. Some tests (obtained with a task used for studying monkeys engaged in learning reaching-action sequences) show that the model is computationally sound and capable of learning to select sensorimotor primitives from the postures’ continuous space on the basis of their population encoding.
When monkeys tackle novel complex behavioral tasks by trial-and-error they select actions from re... more When monkeys tackle novel complex behavioral tasks by trial-and-error they select actions from repertoires of sensorimotor primitives that allow them to search solutions in a space which is coarser than the space of fine movements. Neuroscientific findings suggested that upper-limb sensorimotor primitives might be encoded, in terms of the final goal-postures they pursue, in premotor cortex. A previous work by the authors reproduced these results in a model based on the idea that cortical pathways learn sensorimotor primitives while basal ganglia learn to assemble and trigger them to pursue complex reward-based goals. This paper extends that model in several directions: a) it uses a Kohonen network to create a neural map with population encoding of postural primitives; b) it proposes an actor-critic reinforcement learning algorithm capable of learning to select those primitives in a biologically plausible fashion (i.e., through a dynamic competition between postures); c) it proposes a procedure to pre-train the actor to select promising primitives when tackling novel reinforcement learning tasks. Some tests (obtained with a task used for studying monkeys engaged in learning reaching-action sequences) show that the model is computationally sound and capable of learning to select sensorimotor primitives from the postures’ continuous space on the basis of their population encoding.
In this paper we describe a new approach in Evolutionary Robotics according to which human breede... more In this paper we describe a new approach in Evolutionary Robotics according to which human breeders are involved in the evo- lutionary process. While traditionally robots are selected to reproduce automatically according to a fltness formula, a quantitative and strictly deflned measure, human breeders can operate a selection based on qual- itative criteria, rewarding behaviors that can slip between the meshes woven by the fltness formula. In the authors' opinion this may bring advantages to the Evolutionary Robotics methodology, allowing for the production of robots that display more and more multiform behavior. In order to illustrate this approach, a piece of software, BreedBot, was developed in which human breeders can intervene in evolving robots, complementing the automatic evaluation. After describing the software, some results on sample evolutionary processes are reported showing that the joint use of human and artiflcial selection on an exploration task generates robots ...
Particolarmente originali e attraenti sono i giochi per bambini di tipo “playware”, che prevedono... more Particolarmente originali e attraenti sono i giochi per bambini di tipo “playware”, che prevedono l’uso di tecnologie pervasive volte a favorire lo sviluppo di nuovi scenari ludici dove il giocare affina contemporaneamente le abilità cognitive, motorie e sociali. Il gioco consiste in una corsa tra robot fuori dal comune poiché il robot vincitore viene decretato in base alle azioni che il giocatore compie nel mondo reale.
Robots are attractive for public as they represent a clear and ambitious scientific challenge: an... more Robots are attractive for public as they represent a clear and ambitious scientific challenge: an artificial being creation similar to natural ones by man. The popular enthusiasm grows up in parallel with huge scientific research progress in this field. On this success wave many Science Centers, Science Museums and Science Festivals propose spaces and moments to spread theoretic, methodological and technical issues of this discipline. In this paper we describe some of our exhibits prototypes that were designed and carried out to communicate main approaches of modern robotics (telerobotics, cognitive robotics, autonomous and evolutionary robotics and collective robotics). These exhibits have been conceived as laboratories where a visitor can experiment and put the hands on various kinds of robot. They have been accomplished by putting together cheap and easily available materials. This way it has been possible to present these exhibits in many important scientific divulgation events throughout Europe with a relatively modest expense.
When monkeys tackle novel complex behavioral tasks by trial-and-error they select actions from re... more When monkeys tackle novel complex behavioral tasks by trial-and-error they select actions from repertoires of sensorimotor primitives that allow them to search solutions in a space which is coarser than the space of fine movements. Neuroscientific findings suggested that upper-limb sensorimotor primitives might be encoded, in terms of the final goal-postures they pursue, in premotor cortex. A previous work by the authors reproduced these results in a model based on the idea that cortical pathways learn sensorimotor primitives while basal ganglia learn to assemble and trigger them to pursue complex reward-based goals. This paper extends that model in several directions: a) it uses a Kohonen network to create a neural map with population encoding of postural primitives; b) it proposes an actor-critic reinforcement learning algorithm capable of learning to select those primitives in a biologically plausible fashion (i.e., through a dynamic competition between postures); c) it proposes a procedure to pre-train the actor to select promising primitives when tackling novel reinforcement learning tasks. Some tests (obtained with a task used for studying monkeys engaged in learning reaching-action sequences) show that the model is computationally sound and capable of learning to select sensorimotor primitives from the postures’ continuous space on the basis of their population encoding.
When monkeys tackle novel complex behavioral tasks by trial-and-error they select actions from re... more When monkeys tackle novel complex behavioral tasks by trial-and-error they select actions from repertoires of sensorimotor primitives that allow them to search solutions in a space which is coarser than the space of fine movements. Neuroscientific findings suggested that upper-limb sensorimotor primitives might be encoded, in terms of the final goal-postures they pursue, in premotor cortex. A previous work by the authors reproduced these results in a model based on the idea that cortical pathways learn sensorimotor primitives while basal ganglia learn to assemble and trigger them to pursue complex reward-based goals. This paper extends that model in several directions: a) it uses a Kohonen network to create a neural map with population encoding of postural primitives; b) it proposes an actor-critic reinforcement learning algorithm capable of learning to select those primitives in a biologically plausible fashion (i.e., through a dynamic competition between postures); c) it proposes a procedure to pre-train the actor to select promising primitives when tackling novel reinforcement learning tasks. Some tests (obtained with a task used for studying monkeys engaged in learning reaching-action sequences) show that the model is computationally sound and capable of learning to select sensorimotor primitives from the postures’ continuous space on the basis of their population encoding.
Uploads
Papers by Angelo Rega