TY - GEN
T1 - Attractor representations of language-behavior structure in a recurrent neural network for human-robot interaction
AU - Yamada, Tatsuro
AU - Murata, Shingo
AU - Arie, Hiroaki
AU - Ogata, Tetsuya
PY - 2015/12/11
Y1 - 2015/12/11
N2 - In recent years there has been increased interest in studies that explore integrative learning of language and other modalities by using neural network models. However, for practical application to human-robot interaction, the acquired semantic structure between language and meaning has to be available immediately and repeatably whenever necessary, just as in everyday communication. As a solution to this problem, this study proposes a method in which a recurrent neural network self-organizes cyclic attractors that reflect semantic structure and represent interaction flows in its internal dynamics. To evaluate this method we design a simple task in which a human verbally directs a robot, which responds appropriately. Training the network with training data that represent the interaction series, the cyclic attractors that reflect the semantic structure is self-organized. The network first receives a verbal direction, and its internal state moves according to the first half of the cyclic attractors with branch structures corresponding to semantics. After that, the internal state reaches a potential to generate appropriate behavior. Finally, the internal state moves to the second half and converges on the initial point of the cycle while generating the appropriate behavior. By self-organizing such an internal structure in its forward dynamics, the model achieves immediate and repeatable response to linguistic directions. Furthermore, the network self-organizes a fixed-point attractor, and so able to wait for directions. It can thus repeat the interaction flexibly without explicit turn-taking signs.
AB - In recent years there has been increased interest in studies that explore integrative learning of language and other modalities by using neural network models. However, for practical application to human-robot interaction, the acquired semantic structure between language and meaning has to be available immediately and repeatably whenever necessary, just as in everyday communication. As a solution to this problem, this study proposes a method in which a recurrent neural network self-organizes cyclic attractors that reflect semantic structure and represent interaction flows in its internal dynamics. To evaluate this method we design a simple task in which a human verbally directs a robot, which responds appropriately. Training the network with training data that represent the interaction series, the cyclic attractors that reflect the semantic structure is self-organized. The network first receives a verbal direction, and its internal state moves according to the first half of the cyclic attractors with branch structures corresponding to semantics. After that, the internal state reaches a potential to generate appropriate behavior. Finally, the internal state moves to the second half and converges on the initial point of the cycle while generating the appropriate behavior. By self-organizing such an internal structure in its forward dynamics, the model achieves immediate and repeatable response to linguistic directions. Furthermore, the network self-organizes a fixed-point attractor, and so able to wait for directions. It can thus repeat the interaction flexibly without explicit turn-taking signs.
KW - Hidden Markov models
KW - Neurons
KW - Pragmatics
KW - Robots
KW - Semantics
KW - Training
KW - Training data
UR - http://www.scopus.com/inward/record.url?scp=84958151379&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84958151379&partnerID=8YFLogxK
U2 - 10.1109/IROS.2015.7353968
DO - 10.1109/IROS.2015.7353968
M3 - Conference contribution
AN - SCOPUS:84958151379
T3 - IEEE International Conference on Intelligent Robots and Systems
SP - 4179
EP - 4184
BT - IROS Hamburg 2015 - Conference Digest
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - IEEE/RSJ International Conference on Intelligent Robots and Systems, IROS 2015
Y2 - 28 September 2015 through 2 October 2015
ER -