Naohiro Fukumura

Learning goal-directed sensory-based navigation of a mobile robot

Abstract This paper presents a novel scheme for sensory-based navigation of a mobile robot: the robot is trained to learn a goal-directed task under adequate supervision, utilizing local sensory inputs. Focusing on the topological changes of temporal sensory flow, our scheme constructs a correct mapping from sensory input sequences to the maneuvering outputs through neural adaptation such that a hypothetical vector field that achieves the goal can be generated. The simulation experiments show that a robot, utilizing our scheme, can learn tasks of homing and sequential routing successfully in the work space of a certain geometrical complexity.

A computational model for recognizing objects and planning hand shapes in grasping movements

Abstract To execute grasping movements, the primate brain must solve at least two computational problems (i.e. recognition of objects and planning of prehensile hand shapes). From the viewpoint of computational theory, we hypothesize that the two problems are not separately solved in the brain; instead, they are merged and transformed into the problem of forming an integrated internal representation of visual information and motor information. To demonstrate the computational potential of our hypothesis, we propose a neural network model that integrates visual information and motor information for preshaping a hand in grasping movements. Network operation is divided into a learning phase and an optimization phase. In the learning phase, an internal model that represents the relation between the visual and motor information on grasped objects is acquired by integrating the two sources of information. In the optimization phase, objects are recognized and the most suitable hand shapes for grasping them are obtained by using a relaxation computation of the network. Successful hand prehension demonstrated by our computer simulation supports the computational plausibility of our hypothesis.

Embedding a grammatical description in deterministic chaos: an experiment in recurrent neural learning

We consider the modeling process of a “biological” agent by combining the concepts of neuroinformatics and deterministic chaos. We assume that an agent observes a target process as a stochastic symbolic process, which is restricted by grammatical constraints. Our main hypothesis is that an agent would learn the target model by reconstructing an equivalent quasi-stochastic process on its deterministic neural dynamics. We employed a recurrent neural network (RNN), which is regarded as an adjustable deterministic dynamical system. Then, we conducted an experiment to observe how the RNN learns to reconstruct the target process, represented by a stochastic finite state machine in the simulation. The result revealed the capability of the RNN to evolve, by means of learning, toward chaos, which is able to mimic a targets stochastic process. We precisely analyzed the evolutionary process as well as the internal representation of the neural dynamics obtained. This analysis enabled us to clarify an interesting mechanism of the self-organization of chaos by means of neural learning, and also showed how grammar can be embedded in the evolved deterministic chaos.

Self-organizing internal representation in learning of navigation: a physical experiment by the mobile robot YAMABICO

This paper discusses a novel scheme for sensory-based navigation of a mobile robot. In our previous work ([Tani and Fukumura, 1994], Neural Networks, 7(3), 553-563), we formulated the problem of goal-directed navigation as an embedding problem of dynamical systems: desired trajectories in a task space should be embedded in an adequate sensory-based internal state space so that a unique mapping from the internal state space to the motor command could be established. In the current formulation a recurrent neural network is employed, which shows that an adequate internal state space can be self-organized, through supervised training with sensorimotor sequences. The experiment was conducted using a real mobile robot equipped with a laser range sensor, demonstrating the validity of the presented scheme by working in a noisy real-world environment. Copyright 1996 Elsevier Science Ltd.

Learning goal-directed navigation as attractor dynamics for a sensory motor system. (An experiment by the mobile robot YAMABICO)

This paper describes experimental results based on the authors prior-proposed scheme: learning of sensory-based, goal-directed behavior. The scheme was implemented on the mobile robot YAMABICO and learning of a set of goal-directed navigations were conducted. The experiment assumed that the robot receives no global information such as position nor prior environment model. Instead, the robot was trained to learn adequate maneuvering in the adopted workspace by building a correct mapping between a spatio-temporal sequence of sensory inputs and maneuvering outputs on a neural structure. The experimental results showed that sufficient training generated rigid dynamical structure of a fixed point and limit cycling in the sensory-based state space, which realized robust navigations of homing and cyclic routing even against certain changes of environment as well as miscellaneous noises in the real world.