Evolving Detectors of 2D Patterns on a Simulated CAM-Brain Machine, an Evolvable Hardware Tool for Building a 75 Million Neuron Artificial Brain

摘要

This paper presents some simulation results of the evolution of 2D visual pattern recognizers to be implemented very shortly on real hardware, namely the "CAM-Brain Machine" (CBM), an FPGA based piece of evolvable hardware which implements a genetic algorithm (GA) to evolve a 3D cellular automata (CA) based neural network circuit module, of approximately 1,000 neurons, in about a second, i.e. a compelte run of a GA, with 10,000s of circuit growths and performance evaluations. Up to 65,000 of these modules, each of which is evolved with a humanly specified function, can be downlloaded into a large RAM space, and interconnected according to humanly specified gvdvips -o SPIE-2000 artificial brain architectures. This RAM, containing an artificial brain with up to 75 million neurons, is then updated by the CBM at a rate of 130 billion CA cells per second. Such speeds will enable real time control of robots and hopefully the birth of a new research field that we call "brain building".The first such artificial brain, to be built at STARLAB in 2000 and beyond, will be used to control the behaviors of a life sized kitten robot called "Robokitty". This kitten robot will need 2D pattern recognizers in the visual section of its artificial brain. This paper presents simulation results on the evolvability and generalization properties of such recognizers.is paper explores a single theme- on-board processing is the best avenue to take advantage of advancements in high-performance computing, high-density memories, communications, and re-programmable architecture technologies. The goal is to break away from "no changes after launch" design to a more flexible environment that can take advantage of changing space reuqirements and needs while the space vehicle is "on orbit."