2009 IEEE 8th International Conference on Development and Learning最新文献_第2页

Efficient exploration and learning of whole body kinematics 对全身运动学的有效探索和学习

2009 IEEE 8th International Conference on Development and Learning Pub Date : 2009-06-05 DOI: 10.1109/DEVLRN.2009.5175522

Matthias Rolf, Jochen J. Steil, M. Gienger

{"title":"Efficient exploration and learning of whole body kinematics","authors":"Matthias Rolf, Jochen J. Steil, M. Gienger","doi":"10.1109/DEVLRN.2009.5175522","DOIUrl":"https://doi.org/10.1109/DEVLRN.2009.5175522","url":null,"abstract":"We present a neural network approach to early motor learning. The goal is to explore the needs for boot-strapping the control of hand movements in a biologically plausible learning scenario. The model is applied to the control of hand postures of the humanoid robot ASIMO by means of full upper body movements. For training, we use an efficient online scheme for recurrent reservoir networks consisting of supervised backpropagation-decorrelation output adaptation and an unsupervised intrinsic plasticity reservoir optimization. We demonstrate that the network can acquire accurate inverse models for the highly redundant ASIMO, applying bi-manual target motions and exploiting all upper body degrees of freedom. We show that very few, but highly symmetric training data is sufficient to generate excellent generalization capabilities to untrained target motions. We also succeed in reproducing real motion recorded from a human demonstrator, massively differing from the training data in range and dynamics. The demonstrated generalization capabilities provide a fundamental prerequisite for an autonomous and incremental motor learning in an developmentally plausible way. Our exploration process - though not yet fully autonomous - clearly shows that goal-directed exploration can, in contrast to “babbling” of joints angles, be done very efficiently even for many degrees of freedom and non-linear kinematic configurations as ASIMOs.","PeriodicalId":192225,"journal":{"name":"2009 IEEE 8th International Conference on Development and Learning","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125700385","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}

引用次数: 41

Orientation Plasticity in Visual Cortex of Mice Reared under Single-Orientation Exposure 单方向暴露饲养小鼠视觉皮层的定向可塑性

2009 IEEE 8th International Conference on Development and Learning Pub Date : 2009-06-05 DOI: 10.1109/DEVLRN.2009.5175530

Takamasa Yoshida, Toshiki Tani, Shigeru Tanaka

{"title":"Orientation Plasticity in Visual Cortex of Mice Reared under Single-Orientation Exposure","authors":"Takamasa Yoshida, Toshiki Tani, Shigeru Tanaka","doi":"10.1109/DEVLRN.2009.5175530","DOIUrl":"https://doi.org/10.1109/DEVLRN.2009.5175530","url":null,"abstract":"To examine whether orientation selectivity in the mouse visual cortex can change depending on visual experience, we reared juvenile and adult mice under single-orientation exposure using cylindrical-lens-fitted goggles that extremely elongate visual images vertically. Immediately after goggle rearing, we performed optical imaging of intrinsic signals in the visual cortex of the mouse, while presenting 6 oriented grating stimuli. The distribution of preferred orientations was markedly biased toward the exposed vertical orientation in juvenile goggle-reared mice, whereas the distribution in normally reared mice showed a maximum at horizontal orientation and a minimum at vertical orientation. In contrast, no significant differences in the orientation distribution were found between 1-week goggle-reared and normally reared adult mice. However, in 2- or 3-week goggle-reared adult mice, the relative area maximally responding to the vertical orientation was slightly larger than that in normally reared adult mice whereas the horizontal bias was preserved. The present study demonstrated that postnatal visual experience can modify orientation selectivity in both juvenile and adult mice.","PeriodicalId":192225,"journal":{"name":"2009 IEEE 8th International Conference on Development and Learning","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128814110","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}

引用次数: 3

Laterally connected lobe component analysis: Precision and topography 横向连接瓣成分分析:精度和地形

2009 IEEE 8th International Conference on Development and Learning Pub Date : 2009-06-05 DOI: 10.1109/DEVLRN.2009.5175541

M. Luciw, J. Weng

{"title":"Laterally connected lobe component analysis: Precision and topography","authors":"M. Luciw, J. Weng","doi":"10.1109/DEVLRN.2009.5175541","DOIUrl":"https://doi.org/10.1109/DEVLRN.2009.5175541","url":null,"abstract":"Due to the pressure of evolution, the brains of organisms need to self-organize at different scales during different developmental stages. In early stages, the brain must organize globally (e.g., large cortical areas) to form “smooth” topographic representation that is critical for superior generalization with its limited connections. At later stages, the brain must fine tune its microstructures of representation for “precision” - high-level performance and specialization. But smoothness and precision are two conflicting criteria. The self-organizing map (SOM) mechanisms of self-organization through isotropic updating and other published computational variants have dealt with global to local smoothing and lateral adaptation, but we show in our work that they are insufficient to deliver superior performance. In this paper, we introduce a combination of several mechanisms that, together, address these two conflicting criteria: lateral excitation through adaptive connections, explicit adaptive top-down connections (attention), dually-optimal lobe component analysis (LCA) for synaptic updating, simulated lateral inhibition through winners-take-all, and a developmental schedule that sets the number of winners, which decreases over time. Major performance improvements due to the combination of these mechanisms are shown in the reported experiments.","PeriodicalId":192225,"journal":{"name":"2009 IEEE 8th International Conference on Development and Learning","volume":"41 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128474734","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}

引用次数: 10

Learning from a tutor: Embodied speech acquisition and imitation learning 向导师学习:具身言语习得和模仿学习

2009 IEEE 8th International Conference on Development and Learning Pub Date : 2009-06-05 DOI: 10.1109/DEVLRN.2009.5175543

M. Vaz, H. Brandl, F. Joublin, C. Goerick

引用次数: 9

Learning independent causes in natural images explains the spacevariant oblique effect 学习自然图像中的独立原因解释了空间变斜效应

2009 IEEE 8th International Conference on Development and Learning Pub Date : 2009-06-05 DOI: 10.1109/DEVLRN.2009.5175534

C. Rothkopf, Thomas H. Weisswange, J. Triesch

{"title":"Learning independent causes in natural images explains the spacevariant oblique effect","authors":"C. Rothkopf, Thomas H. Weisswange, J. Triesch","doi":"10.1109/DEVLRN.2009.5175534","DOIUrl":"https://doi.org/10.1109/DEVLRN.2009.5175534","url":null,"abstract":"The efficient coding hypothesis posits that sensory processing increases independence between neural responses to natural stimuli by removing their statistical redundancy reflective of the structure present in the natural environment. While there is consensus on the role of the statistical structure of the physical environment in shaping the natural input to the sensory system, it is not well understood how the sensory apparatus itself and its active use during behavior determine the statistics of the input. To explore this issue, a virtual human agent is simulated navigating through a wooded environment under full control of its gaze allocation during walking. Independent causes for the images obtained during navigation are learned with algorithms that have been shown to extract computationally useful representations similar to those encountered in the primary visual cortex of the mammalian brain. The distributions of properties of the learned simple cell like units are in good agreement with a wealth of data on the visual system including the oblique effect, the meridional effect, properties of neurons in the macaque visual cortex, and functional Magnetic Resonance Imaging (fMRI) data on orientation selectivity in humans and monkeys. Finally, this analysis sheds new light on the discussion on orientation anisotropies based on carpented environments. Thus, when learning computational representations it is not sufficient to consider only the regularities of the environment but also the regularities imposed by the sensory apparatus and its use during behavior need to be taken into account.","PeriodicalId":192225,"journal":{"name":"2009 IEEE 8th International Conference on Development and Learning","volume":"51 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122092892","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}

引用次数: 24

Human-Robot Interaction in Concept Acquisition: a computational model 概念习得中的人机交互:一个计算模型

2009 IEEE 8th International Conference on Development and Learning Pub Date : 2009-06-05 DOI: 10.1109/DEVLRN.2009.5175532

Joachim de Greeff, F. Delaunay, Tony Belpaeme

引用次数: 26

Learning motor dependent Crutchfield's information distance to anticipate changes in the topology of sensory body maps 学习运动依赖的克兰驰菲尔德信息距离以预测感觉体图拓扑结构的变化

2009 IEEE 8th International Conference on Development and Learning Pub Date : 2009-06-05 DOI: 10.1109/DEVLRN.2009.5175526

Thomas Schatz, Pierre-Yves Oudeyer

引用次数: 10

A theory of architecture for spatial abstraction 空间抽象的建筑理论

2009 IEEE 8th International Conference on Development and Learning Pub Date : 2009-06-05 DOI: 10.1109/DEVLRN.2009.5175545

J. Weng

{"title":"A theory of architecture for spatial abstraction","authors":"J. Weng","doi":"10.1109/DEVLRN.2009.5175545","DOIUrl":"https://doi.org/10.1109/DEVLRN.2009.5175545","url":null,"abstract":"A great mystery is how the brain abstracts during the process of development. It is also unclear how motor actions alter cortical representation. The architecture theory introduced here indicates that for each cortical area, the bottom-up space and top-down space are two sources of its representation — bridge representation which embeds manifolds of both spaces into a single space. A bridge representation has the following properties (a) responses from developed neurons are relatively less sensitive to irrelevant sensory information (i.e., invariants) but are relatively more sensitive to relevant sensory information for classification (i.e., discriminants), (b) neurons form topographic cortical areas according to abstract classes. Both properties transform meaningless (iconic, pixel like) raw sensory inputs into an internal representation with abstract meanings. The most abstract area can be considered as frontal cortex (or motor area if each firing pattern of the motor represents a unique abstract class). Such a cortical representation system is neither a purely symbolic system nor a monolithic meaning system, but is iconic-abstract two-way: bottom-up attention, top-down attention and recognition are all tightly integrated and highly distributed throughout the developmental network.","PeriodicalId":192225,"journal":{"name":"2009 IEEE 8th International Conference on Development and Learning","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129413699","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}

引用次数: 3

Can reinforcement learning explain the development of causal inference in multisensory integration? 强化学习能否解释多感觉统合中因果推理的发展?

2009 IEEE 8th International Conference on Development and Learning Pub Date : 2009-06-05 DOI: 10.1109/DEVLRN.2009.5175531

Thomas H. Weisswange, C. Rothkopf, Tobias Rodemann, J. Triesch

{"title":"Can reinforcement learning explain the development of causal inference in multisensory integration?","authors":"Thomas H. Weisswange, C. Rothkopf, Tobias Rodemann, J. Triesch","doi":"10.1109/DEVLRN.2009.5175531","DOIUrl":"https://doi.org/10.1109/DEVLRN.2009.5175531","url":null,"abstract":"Bayesian inference techniques have been used to understand the performance of human subjects on a large number of sensory tasks. Particularly, it has been shown that humans integrate sensory inputs from multiple cues in an optimal way in many conditions. Recently it has also been proposed that causal inference [1] can well describe the way humans select the most plausible model for a given input. It is still unclear how those problems are solved in the brain. Also, considering that infants do not yet behave as ideal observers [2]–[4], it is interesting to ask how the related abilities can develop. We present a reinforcement learning approach to this problem. An orienting task is used in which we reward the model for a correct movement to the origin of noisy audio visual signals. We show that the model learns to do cue-integration and model selection, in this case inferring the number of objects. Its behaviour also includes differences in reliability between the two modalities. All of that comes without any prior knowledge by simple interaction with the environment.","PeriodicalId":192225,"journal":{"name":"2009 IEEE 8th International Conference on Development and Learning","volume":"59 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2009-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126440859","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}

引用次数: 7

Robotic hand-eye coordination without global reference: A biologically inspired learning scheme 没有全局参考的机器人手眼协调:生物学启发的学习方案

2009 IEEE 8th International Conference on Development and Learning Pub Date : 2009-06-05 DOI: 10.1109/DEVLRN.2009.5175514

Martin Hulse, Sebastian McBrid, Mark H. Lee

引用次数: 21