Minimalist mobile robotics: a colony-style architecture for an artificial creature
Minimalist mobile robotics: a colony-style architecture for an artificial creature
ALVINN: an autonomous land vehicle in a neural network
Advances in neural information processing systems 1
Introduction to the theory of neural computation
Introduction to the theory of neural computation
Parallel distributed processing: explorations in the microstructure of cognition, vol. 1
Advances in neural information processing systems 2
Maximum likelihood competitive learning
Advances in neural information processing systems 2
Mapbuilding using self-organising networks in “really useful robots”
Proceedings of the first international conference on simulation of adaptive behavior on From animals to animats
Learning hill-climbing functions as a strategy for generating behaviors in a mobile robot
Proceedings of the first international conference on simulation of adaptive behavior on From animals to animats
Learning to Perceive and Act by Trial and Error
Machine Learning
Learning in embedded systems
Evolving visually guided robots
Proceedings of the second international conference on From animals to animats 2 : simulation of adaptive behavior: simulation of adaptive behavior
Evolving dynamical neural networks for adaptive behavior
Adaptive Behavior
Sparse distributed memory and related models
Associative neural memories
An active vision architecture based on iconic representations
Artificial Intelligence - Special volume on computer vision
Using emergent modularity to develop control systems for mobile robots
Adaptive Behavior - Special issue on environment structure and behavior
Sparse Distributed Memory
Advances in Neural Information Processing Systems 5, [NIPS Conference]
Object indexing using an iconic sparse distributed memory
ICCV '95 Proceedings of the Fifth International Conference on Computer Vision
Reinforcement learning: a survey
Journal of Artificial Intelligence Research
Natural basis functions and topographic memory for face recognition
IJCAI'95 Proceedings of the 14th international joint conference on Artificial intelligence - Volume 1
Hidden state and reinforcement learning with instance-based stateidentification
IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics
| Hi-index | 0.00 |
We describe a general framework for learningperception-based navigational behaviors in autonomous mobile robots.A hierarchical behavior-based decomposition of the controlarchitecture is used to facilitate efficient modular learning. Lowerlevel reactive behaviors such as collision detection and obstacleavoidance are learned using a stochastic hill-climbing method whilehigher level goal-directed navigation is achieved using aself-organizing sparse distributed memory. The memory is initiallytrained by teleoperating the robot on a small number of paths withina given domain of interest. During training, the vectors in thesensory space as well as the motor space are continually adaptedusing a form of competitive learning to yield basis vectors thatefficiently span the sensorimotor space. After training, the robotnavigates from arbitrary locations to a desired goal location usingmotor output vectors computed by a saliency-based weighted averagingscheme. The pervasive problem of perceptual aliasing in finite-orderMarkovian environments is handled by allowing both current as wellas the set of immediately preceding perceptual inputs to predict themotor output vector for the current time instant. We describeexperimental and simulation results obtained using a mobile robotequipped with bump sensors, photosensors and infrared receivers,navigating within an enclosed obstacle-ridden arena. The resultsindicate that the method performs successfully in a number ofnavigational tasks exhibiting varying degrees of perceptualaliasing.