Acquiring Localization Ability by Self-organization in a Non-linear Motor System
Hiromichi Nakashima, Tetsuya Matsumoto, Noboru Ohnishi · IEEJ Transactions on Electronics Information and Systems · 2002
We propose a learning model of sound source localization in a non-linear motor system with self-organizing through the interaction between motion and audio-visual sensing. This ability is acquired by repeated sensing and moving. The model consists of two modules -a visual estimation module and an auditory estimation module. These modules consist of a self-organizing feature map. These two modules learn at once. The visual estimation module learns by direct inverse modeling and the auditory estimation module uses the output from the visual estimation module to learn. We conducted computer simulation experiments to investigate the validity of the proposed module. The experimental results demonstrate that sound source localization ability can be acquired without supervision for a nonlinear system and the system is robust for noise and experimental environment.