Pub Date : 2015-12-07DOI: 10.1109/DEVLRN.2015.7346113
Stephan K. U. Zibner, Jan Tekülve, G. Schöner
We present a neuro-dynamic model of looking, reaching, and grasping movements in infants in three pre-reaching phases. We attribute the evolution from pre-reaches to their suppression and subsequent re-emergence reported in a longitudinal study of von Hofsten [1] to the development of the sequential organization of movements, through which a set of elementary movements (visual fixation, reaching, opening the hand) are coordinated in time. The spatial precision hypothesis, which has emerged from work on spatial, visual, and action working memory, characterizes developmental changes as a change from strongly input-driven to more strongly interaction-dominated neural dynamics. Applying this hypothesis to reaching, we propose that the intention to reach is increasingly able to suppress competing movement behaviors, enabling object-oriented reaches. We evaluate three versions of the model that capture the three phases reported by von Hofsten and illustrate the properties of the movement model in simulations and in demonstration on a NAO robot.
{"title":"The sequential organization of movement is critical to the development of reaching: A neural dynamics account","authors":"Stephan K. U. Zibner, Jan Tekülve, G. Schöner","doi":"10.1109/DEVLRN.2015.7346113","DOIUrl":"https://doi.org/10.1109/DEVLRN.2015.7346113","url":null,"abstract":"We present a neuro-dynamic model of looking, reaching, and grasping movements in infants in three pre-reaching phases. We attribute the evolution from pre-reaches to their suppression and subsequent re-emergence reported in a longitudinal study of von Hofsten [1] to the development of the sequential organization of movements, through which a set of elementary movements (visual fixation, reaching, opening the hand) are coordinated in time. The spatial precision hypothesis, which has emerged from work on spatial, visual, and action working memory, characterizes developmental changes as a change from strongly input-driven to more strongly interaction-dominated neural dynamics. Applying this hypothesis to reaching, we propose that the intention to reach is increasingly able to suppress competing movement behaviors, enabling object-oriented reaches. We evaluate three versions of the model that capture the three phases reported by von Hofsten and illustrate the properties of the movement model in simulations and in demonstration on a NAO robot.","PeriodicalId":164756,"journal":{"name":"2015 Joint IEEE International Conference on Development and Learning and Epigenetic Robotics (ICDL-EpiRob)","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-12-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114135066","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2015-12-07DOI: 10.1109/DEVLRN.2015.7346122
T. Chandrapala, Bertram E. Shi, J. Triesch
Neural development in the visual cortex depends on the visual experience during the so-called critical period. Recent experiments have shown that under normal conditions rodents develop binocular receptive fields which have similar orientation preferences for the left and right eyes. In contrast, under conditions of monocular deprivation during the critical period, this orientation alignment does not happen. Here we propose a computational model to explain the process of orientation alignment, its underlying mechanisms, and its failure in case of monocular deprivation or uncorrelated binocular inputs. Our model is based on the recently proposed Active Efficient Coding framework that jointly develops eye movement control and sensory representations. Our model suggests that the active maintenance of a binocular visual field, which leads to correlated visual inputs from the two eyes, is essential for the process of orientation alignment. This behavior is analogous to vergence control in primates. However, due to the fact that rodents have large receptive fields with low spatial frequency tuning, the coordination of the eyes need not be very precise. The model also suggests that it is not necessary that coordinated binocular vision be maintained continuously in order for orientation alignment to develop.
{"title":"Active maintenance of binocular correspondence leads to orientation alignment of visual receptive fields","authors":"T. Chandrapala, Bertram E. Shi, J. Triesch","doi":"10.1109/DEVLRN.2015.7346122","DOIUrl":"https://doi.org/10.1109/DEVLRN.2015.7346122","url":null,"abstract":"Neural development in the visual cortex depends on the visual experience during the so-called critical period. Recent experiments have shown that under normal conditions rodents develop binocular receptive fields which have similar orientation preferences for the left and right eyes. In contrast, under conditions of monocular deprivation during the critical period, this orientation alignment does not happen. Here we propose a computational model to explain the process of orientation alignment, its underlying mechanisms, and its failure in case of monocular deprivation or uncorrelated binocular inputs. Our model is based on the recently proposed Active Efficient Coding framework that jointly develops eye movement control and sensory representations. Our model suggests that the active maintenance of a binocular visual field, which leads to correlated visual inputs from the two eyes, is essential for the process of orientation alignment. This behavior is analogous to vergence control in primates. However, due to the fact that rodents have large receptive fields with low spatial frequency tuning, the coordination of the eyes need not be very precise. The model also suggests that it is not necessary that coordinated binocular vision be maintained continuously in order for orientation alignment to develop.","PeriodicalId":164756,"journal":{"name":"2015 Joint IEEE International Conference on Development and Learning and Epigenetic Robotics (ICDL-EpiRob)","volume":"121 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-12-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123054292","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2015-12-07DOI: 10.1109/DEVLRN.2015.7346135
M. Schlesinger, Scott P. Johnson, Dima Amso
We recently demonstrated that the gaze sequences produced by infants during an habituation event predict their looking times to both the habituation and (one of two) posthabituation test events [1]. Specifically, we trained a simple recurrent network (SRN) to predict infants' habituation gaze sequences. Sequences that were easier for the SRN to learn were associated with shorter looking times at the end of habituation, as well as longer looking times to one of two posthabituation test events. In the current study, we extended these findings by applying the sequential-learnability model to a new set of looking-time data, in which an important visual cue was removed from the habituation and test events. Following our previous work, we predicted that “learnability” of infants' habituation gaze sequences would predict their habituation looking time. However, unlike the previous study, we also predicted that habituation gaze sequences would not predict looking time to either of the posthabituation test events. The results were consistent with both of these predictions.
{"title":"Do infants' gaze sequences predict their looking time? Testing the sequential-learnability model","authors":"M. Schlesinger, Scott P. Johnson, Dima Amso","doi":"10.1109/DEVLRN.2015.7346135","DOIUrl":"https://doi.org/10.1109/DEVLRN.2015.7346135","url":null,"abstract":"We recently demonstrated that the gaze sequences produced by infants during an habituation event predict their looking times to both the habituation and (one of two) posthabituation test events [1]. Specifically, we trained a simple recurrent network (SRN) to predict infants' habituation gaze sequences. Sequences that were easier for the SRN to learn were associated with shorter looking times at the end of habituation, as well as longer looking times to one of two posthabituation test events. In the current study, we extended these findings by applying the sequential-learnability model to a new set of looking-time data, in which an important visual cue was removed from the habituation and test events. Following our previous work, we predicted that “learnability” of infants' habituation gaze sequences would predict their habituation looking time. However, unlike the previous study, we also predicted that habituation gaze sequences would not predict looking time to either of the posthabituation test events. The results were consistent with both of these predictions.","PeriodicalId":164756,"journal":{"name":"2015 Joint IEEE International Conference on Development and Learning and Epigenetic Robotics (ICDL-EpiRob)","volume":"103 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-12-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123348146","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2015-12-07DOI: 10.1109/DEVLRN.2015.7346152
Erwan Renaudo, Benoît Girard, R. Chatila, M. Khamassi
Research in the fields of Psychology and Neuroscience have provided strong evidence that mammals can adaptively switch between goal-directed behaviors - i.e. deliberative decisions based on costly but flexible planned long-term consequences of actions - and habitual behaviors - i.e. reactive behaviors that are efficient when the environment is stable but inflexible in the case of environmental changes. However, the computational principles underlying this switching ability are not yet understood, and several alternative criteria have been proposed, each tested on specific subsets of experimental datasets. Here we present a neurorobotic implementation and comparison of such type of criteria, plus some new ones imported from the field of ensemble reinforcement learning, with a twofold objective: on the one hand exploring the possible efficiency of such bio-inspired principles to enable robots to have more behavioral flexibility during autonomous development and learning; on the other hand, analyzing whether an asynchronous continuous robotic simulation and comparison of these criteria in a common task can feed current debates in the Psychological and Neuroscience fields. We evaluate these methods in an apparently simple repetitive cube-pushing task on a simulated conveyor belt, but which imposes to the robot constant trade-offs between speed and accuracy and between stability and abrupt changes. Our results show that if overall performance is not improved by using multiple behavioral systems in a stable environment, these methods allow for a better adaptation to environmental changes. The Voting methods and Boltzmann addition, from ensemble reinforcement learning, give the best performance, providing an interesting alternative to Expert selection.
{"title":"Which criteria for autonomously shifting between goal-directed and habitual behaviors in robots?","authors":"Erwan Renaudo, Benoît Girard, R. Chatila, M. Khamassi","doi":"10.1109/DEVLRN.2015.7346152","DOIUrl":"https://doi.org/10.1109/DEVLRN.2015.7346152","url":null,"abstract":"Research in the fields of Psychology and Neuroscience have provided strong evidence that mammals can adaptively switch between goal-directed behaviors - i.e. deliberative decisions based on costly but flexible planned long-term consequences of actions - and habitual behaviors - i.e. reactive behaviors that are efficient when the environment is stable but inflexible in the case of environmental changes. However, the computational principles underlying this switching ability are not yet understood, and several alternative criteria have been proposed, each tested on specific subsets of experimental datasets. Here we present a neurorobotic implementation and comparison of such type of criteria, plus some new ones imported from the field of ensemble reinforcement learning, with a twofold objective: on the one hand exploring the possible efficiency of such bio-inspired principles to enable robots to have more behavioral flexibility during autonomous development and learning; on the other hand, analyzing whether an asynchronous continuous robotic simulation and comparison of these criteria in a common task can feed current debates in the Psychological and Neuroscience fields. We evaluate these methods in an apparently simple repetitive cube-pushing task on a simulated conveyor belt, but which imposes to the robot constant trade-offs between speed and accuracy and between stability and abrupt changes. Our results show that if overall performance is not improved by using multiple behavioral systems in a stable environment, these methods allow for a better adaptation to environmental changes. The Voting methods and Boltzmann addition, from ensemble reinforcement learning, give the best performance, providing an interesting alternative to Expert selection.","PeriodicalId":164756,"journal":{"name":"2015 Joint IEEE International Conference on Development and Learning and Epigenetic Robotics (ICDL-EpiRob)","volume":"15 4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-12-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131133233","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2015-12-07DOI: 10.1109/DEVLRN.2015.7346136
Mathias Schmerling, G. Schillaci, V. Hafner
Visuo-motor coordination is known to be highly important for the development of a broader range of cognitive and motor skills in human infants and can thus be considered one of the key skills for robots to master. In this paper, we investigate how a recent concept in developmental robotics, referred to as goal babbling, relates to a visuo-motor coordination task in the humanoid robot Aldebaran Nao that requires coordinated control of two subsystems of motors, namely head and arm motors. The idea of goal babbling builds on findings in developmental psychology showing that human infants attempt goal-directed movements early on in their development enabling them to rapidly and efficiently bootstrap their motor system. Goal babbling has been shown to be superior to the classical idea of random motor babbling for the learning of body kinematics in robotic systems, in particular for systems with many degrees of freedom. Our results not only support the utility of goal babbling for the acquisition of visuo-motor coordination skills but also suggest that goal babbling is particularly effective in the case where two separate motor sub-systems, head and arm, need to be coordinated.
{"title":"Goal-directed learning of hand-eye coordination in a humanoid robot","authors":"Mathias Schmerling, G. Schillaci, V. Hafner","doi":"10.1109/DEVLRN.2015.7346136","DOIUrl":"https://doi.org/10.1109/DEVLRN.2015.7346136","url":null,"abstract":"Visuo-motor coordination is known to be highly important for the development of a broader range of cognitive and motor skills in human infants and can thus be considered one of the key skills for robots to master. In this paper, we investigate how a recent concept in developmental robotics, referred to as goal babbling, relates to a visuo-motor coordination task in the humanoid robot Aldebaran Nao that requires coordinated control of two subsystems of motors, namely head and arm motors. The idea of goal babbling builds on findings in developmental psychology showing that human infants attempt goal-directed movements early on in their development enabling them to rapidly and efficiently bootstrap their motor system. Goal babbling has been shown to be superior to the classical idea of random motor babbling for the learning of body kinematics in robotic systems, in particular for systems with many degrees of freedom. Our results not only support the utility of goal babbling for the acquisition of visuo-motor coordination skills but also suggest that goal babbling is particularly effective in the case where two separate motor sub-systems, head and arm, need to be coordinated.","PeriodicalId":164756,"journal":{"name":"2015 Joint IEEE International Conference on Development and Learning and Epigenetic Robotics (ICDL-EpiRob)","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-12-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133174419","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2015-12-07DOI: 10.1109/DEVLRN.2015.7346121
Iris Oved, Shaun Nichols, D. Barner
Many cognitive scientists take it for granted that concepts like CAT (mental terms that are expressed with single nouns) can be learned by observing a co-occurrence in superficial properties, such as having fur, being 4-legged, and tending to purr, and then building a complex category representation from representations for those superficial properties. A less popular account, known as Psychological Essentialism, claims that concepts like CAT pick out deep, hidden properties (essences) that are causal explanations for observable co-occurrences in superficial properties. The trouble is, Psychological Essentialism lacks an account of how such essentialist concepts could be learned, and often adopt the unpalatable conclusion that such concepts are innate. Developmental roboticists have recently started implementing systems that employ learned hidden/latent variables. The present paper spells out a learning theory for essentialist concepts, and presents two psychology experiments that help support the account over the associationist alternative.
{"title":"A learning model for essentialist concepts","authors":"Iris Oved, Shaun Nichols, D. Barner","doi":"10.1109/DEVLRN.2015.7346121","DOIUrl":"https://doi.org/10.1109/DEVLRN.2015.7346121","url":null,"abstract":"Many cognitive scientists take it for granted that concepts like CAT (mental terms that are expressed with single nouns) can be learned by observing a co-occurrence in superficial properties, such as having fur, being 4-legged, and tending to purr, and then building a complex category representation from representations for those superficial properties. A less popular account, known as Psychological Essentialism, claims that concepts like CAT pick out deep, hidden properties (essences) that are causal explanations for observable co-occurrences in superficial properties. The trouble is, Psychological Essentialism lacks an account of how such essentialist concepts could be learned, and often adopt the unpalatable conclusion that such concepts are innate. Developmental roboticists have recently started implementing systems that employ learned hidden/latent variables. The present paper spells out a learning theory for essentialist concepts, and presents two psychology experiments that help support the account over the associationist alternative.","PeriodicalId":164756,"journal":{"name":"2015 Joint IEEE International Conference on Development and Learning and Epigenetic Robotics (ICDL-EpiRob)","volume":"64 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-12-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115212977","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2015-12-07DOI: 10.1109/DEVLRN.2015.7346127
Suhas E. Chelian, Jaehyon Paik, P. Pirolli, C. Lebiere, Rajan Bhattacharyya
Procedural memory and episodic memory are known to be distinct and both underlie the performance of many tasks. Reinforcement learning (RL) and instance-based learning (IBL) represent common approaches to modeling procedural and episodic memory in that order. In this work, we present a neural model utilizing RL dynamics and an ACT-R model utilizing IBL productions to the task of modeling human decision making in a prognostic foraging task. The task performed was derived from a geospatial intelligence domain wherein agents must choose among information sources to more accurately predict the actions of an adversary. Results from both models are compared to human data and suggest that information gain is an important component in modeling decision-making behavior using either memory system; with respect to the episodic memory approach, the procedural memory approach has a small but significant advantage in fitting human data. Finally, we discuss the interactions of multi-memory systems in complex decision-making tasks.
{"title":"Reinforcement learning and instance-based learning approaches to modeling human decision making in a prognostic foraging task","authors":"Suhas E. Chelian, Jaehyon Paik, P. Pirolli, C. Lebiere, Rajan Bhattacharyya","doi":"10.1109/DEVLRN.2015.7346127","DOIUrl":"https://doi.org/10.1109/DEVLRN.2015.7346127","url":null,"abstract":"Procedural memory and episodic memory are known to be distinct and both underlie the performance of many tasks. Reinforcement learning (RL) and instance-based learning (IBL) represent common approaches to modeling procedural and episodic memory in that order. In this work, we present a neural model utilizing RL dynamics and an ACT-R model utilizing IBL productions to the task of modeling human decision making in a prognostic foraging task. The task performed was derived from a geospatial intelligence domain wherein agents must choose among information sources to more accurately predict the actions of an adversary. Results from both models are compared to human data and suggest that information gain is an important component in modeling decision-making behavior using either memory system; with respect to the episodic memory approach, the procedural memory approach has a small but significant advantage in fitting human data. Finally, we discuss the interactions of multi-memory systems in complex decision-making tasks.","PeriodicalId":164756,"journal":{"name":"2015 Joint IEEE International Conference on Development and Learning and Epigenetic Robotics (ICDL-EpiRob)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-12-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123920047","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2015-12-07DOI: 10.1109/DEVLRN.2015.7346139
Clément Moulin-Frier, Martí Sánchez-Fibla, P. Verschure
We provide a computational model showing how turn-taking behaviors can self-organize out of sensorimotor interactions between vocalizing agents. Recent hypotheses propose that turn-taking behaviors in certain primate species emerge from a need to maintain vocal contact in a group (e.g. in dense environments preventing visual contact). In this context, vocalizations can convey information about the presence of each group member and taking turns allow to minimize the vocal signal interferences. We consider agents equipped with a cognitive architecture based on two coupled control loops: a reactive one implementing a basic regulatory behavior to maintain vocal listening and an adaptive one learning an action policy to maximize vocal contact among group members. We show that the reactive process bootstraps the adaptive learning to converge toward a collective turn-taking strategy. This model provides a computational support to the hypothesis that turn-taking can emerge from functional constraints related to group cohesion and inter-individual vocal signal interferences. We suggest future directions of research to understand how social behaviors can result from sensorimotor interactions.
{"title":"Autonomous development of turn-taking behaviors in agent populations: A computational study","authors":"Clément Moulin-Frier, Martí Sánchez-Fibla, P. Verschure","doi":"10.1109/DEVLRN.2015.7346139","DOIUrl":"https://doi.org/10.1109/DEVLRN.2015.7346139","url":null,"abstract":"We provide a computational model showing how turn-taking behaviors can self-organize out of sensorimotor interactions between vocalizing agents. Recent hypotheses propose that turn-taking behaviors in certain primate species emerge from a need to maintain vocal contact in a group (e.g. in dense environments preventing visual contact). In this context, vocalizations can convey information about the presence of each group member and taking turns allow to minimize the vocal signal interferences. We consider agents equipped with a cognitive architecture based on two coupled control loops: a reactive one implementing a basic regulatory behavior to maintain vocal listening and an adaptive one learning an action policy to maximize vocal contact among group members. We show that the reactive process bootstraps the adaptive learning to converge toward a collective turn-taking strategy. This model provides a computational support to the hypothesis that turn-taking can emerge from functional constraints related to group cohesion and inter-individual vocal signal interferences. We suggest future directions of research to understand how social behaviors can result from sensorimotor interactions.","PeriodicalId":164756,"journal":{"name":"2015 Joint IEEE International Conference on Development and Learning and Epigenetic Robotics (ICDL-EpiRob)","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-12-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116433595","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2015-12-07DOI: 10.1109/DEVLRN.2015.7346124
Ali Mahdi, M. Schlesinger, Dima Amso, Jun Qin
Statistical relationships between human vision and natural scene statistics have been addressed. Contrast is fundamental image property. We proposed that infant visual exploration and learning development could be analyzed using contrast entropy minimization. Ten infants viewed 16 naturalistic images. Fixations of the infants were recorded. Infant gaze patterns are compared between individual infants. Infants are compared to the fixations predicted by the contrast entropy minimization. The results show that infant tends to have different gaze patterns, and infant entropy drops gradually as fixations are made.
{"title":"Infants gaze pattern analyzing using contrast entropy minimization","authors":"Ali Mahdi, M. Schlesinger, Dima Amso, Jun Qin","doi":"10.1109/DEVLRN.2015.7346124","DOIUrl":"https://doi.org/10.1109/DEVLRN.2015.7346124","url":null,"abstract":"Statistical relationships between human vision and natural scene statistics have been addressed. Contrast is fundamental image property. We proposed that infant visual exploration and learning development could be analyzed using contrast entropy minimization. Ten infants viewed 16 naturalistic images. Fixations of the infants were recorded. Infant gaze patterns are compared between individual infants. Infants are compared to the fixations predicted by the contrast entropy minimization. The results show that infant tends to have different gaze patterns, and infant entropy drops gradually as fixations are made.","PeriodicalId":164756,"journal":{"name":"2015 Joint IEEE International Conference on Development and Learning and Epigenetic Robotics (ICDL-EpiRob)","volume":"100 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-12-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124097014","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2015-12-07DOI: 10.1109/DEVLRN.2015.7346115
Kenichi Narioka, Jochen J. Steil
The development of skills in human infants sometimes proceeds so-called U-shaped form, an sequence of disappearance and reappearance of a skill, which has been reported in [1] for pre-reaching to balls presented to infants of early age. We use Goal Babbling, a computational model for exploratory learning of motor skills, to model this U-shaped learning dynamics and furthermore also the qualitative differences caused by presentation modes of the goals. To this aim, we introduce developmentally plausible motor noise and adaptive learning rates in Goal Babbling and show through extensive simulation that U-shaped motor development emerges, thereby reproducing the findings of [1] in our computational model. The results suggest that the disappearance of the skill is caused by a combination of improving the skill and reducing the motor noise in the course of the maturing process, while the reappearance of the skill and its precision is directly reflecting the progress in learning the motor coordination.
{"title":"U-shaped motor development emerges from Goal Babbling with intrinsic motor noise","authors":"Kenichi Narioka, Jochen J. Steil","doi":"10.1109/DEVLRN.2015.7346115","DOIUrl":"https://doi.org/10.1109/DEVLRN.2015.7346115","url":null,"abstract":"The development of skills in human infants sometimes proceeds so-called U-shaped form, an sequence of disappearance and reappearance of a skill, which has been reported in [1] for pre-reaching to balls presented to infants of early age. We use Goal Babbling, a computational model for exploratory learning of motor skills, to model this U-shaped learning dynamics and furthermore also the qualitative differences caused by presentation modes of the goals. To this aim, we introduce developmentally plausible motor noise and adaptive learning rates in Goal Babbling and show through extensive simulation that U-shaped motor development emerges, thereby reproducing the findings of [1] in our computational model. The results suggest that the disappearance of the skill is caused by a combination of improving the skill and reducing the motor noise in the course of the maturing process, while the reappearance of the skill and its precision is directly reflecting the progress in learning the motor coordination.","PeriodicalId":164756,"journal":{"name":"2015 Joint IEEE International Conference on Development and Learning and Epigenetic Robotics (ICDL-EpiRob)","volume":"105 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-12-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126705100","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：