Artificial neural networks achieve fast parallel processing via massively parallel non-linear com... more Artificial neural networks achieve fast parallel processing via massively parallel non-linear computational elements. Most neural network models base their ability to adapt to problems on changing the strength of the interconnections between computational elements according to a given learning algorithm. However, constrained interconnection structures may limit such ability. Field programmable hardware devices allow the implementation of neural networks with in-circuit structure adaptation. This paper describes an FPGA implementation of the FAST (Flexible Adaptable-Size Topology) architecture, a neural network that dynamically changes its size. Since initial experiments indicated a good performance on pattern clustering tasks, we have applied our dynamic-structure FAST neural network to an image segmentation and recognition problem.
Living beings are complex systems exhibiting a range of desirable qualifications that have eluded... more Living beings are complex systems exhibiting a range of desirable qualifications that have eluded realization by traditional engineering methodologies. In recent years we are witness to a growing interest in Nature exhibited by engineers, wishing to imitate the observed processes, thereby creating powerful problem-solving methodologies. If one considers Life on earth since its very beginning, three levels of organization can be distinguished: the phylogenetic level concerns the temporal evolution of the genetic programs within individuals and species, the ontogenetic level concerns the developmental process of a single multicellular organism, and the epigenetic level concerns the learning processes during an individual organism's lifetime. In analogy to Nature, the space of bioinspired systems can be partitioned along these three axes, phylogeny, ontogeny, and epigenesis, giving rise to the POE model. This paper is an exposition and examination of bio-inspired systems within the POE framework. We first discuss each of the three axes separately, considering the systems created to date and plotting directions for continued progress along the axis in question. We end our exposition by a discussion of possible research directions, involving the construction of bio-inspired systems that are situated along two, and ultimately all three axes. This presents a vision for the future which will see the advent of novel systems, inspired by the powerful examples provided by Nature.
Results of the URSEIS &am... more Results of the URSEIS '95 integrated seismic experiment document the lithospheric structure of an intact Paleozoic collisional orogen in the Ural Mountains. Hybrid-source seismic reflection and refraction data provide images of a crustal-scale collisional fabric and a pronounced crustal root preserved since Paleozoic time. Mantle reflections are observed at depths of more than 150 kilometers, possibly representing the base of
Artificial neural networks achieve fast parallel processing via massively parallel non-linear com... more Artificial neural networks achieve fast parallel processing via massively parallel non-linear computational elements. Most neural network models base their ability to adapt to problems on changing the strength of the interconnections between computational elements according to a given learning algorithm. However, constrained interconnection structures may limit such ability. Field programmable hardware devices allow the implementation of neural networks with in-circuit structure adaptation. This paper describes an FPGA implementation of the FAST (Flexible Adaptable-Size Topology) architecture, a neural network that dynamically changes its size. Since initial experiments indicated a good performance on pattern clustering tasks, we have applied our dynamic-structure FAST neural network to an image segmentation and recognition problem.
Living beings are complex systems exhibiting a range of desirable qualifications that have eluded... more Living beings are complex systems exhibiting a range of desirable qualifications that have eluded realization by traditional engineering methodologies. In recent years we are witness to a growing interest in Nature exhibited by engineers, wishing to imitate the observed processes, thereby creating powerful problem-solving methodologies. If one considers Life on earth since its very beginning, three levels of organization can be distinguished: the phylogenetic level concerns the temporal evolution of the genetic programs within individuals and species, the ontogenetic level concerns the developmental process of a single multicellular organism, and the epigenetic level concerns the learning processes during an individual organism's lifetime. In analogy to Nature, the space of bioinspired systems can be partitioned along these three axes, phylogeny, ontogeny, and epigenesis, giving rise to the POE model. This paper is an exposition and examination of bio-inspired systems within the POE framework. We first discuss each of the three axes separately, considering the systems created to date and plotting directions for continued progress along the axis in question. We end our exposition by a discussion of possible research directions, involving the construction of bio-inspired systems that are situated along two, and ultimately all three axes. This presents a vision for the future which will see the advent of novel systems, inspired by the powerful examples provided by Nature.
Results of the URSEIS &am... more Results of the URSEIS '95 integrated seismic experiment document the lithospheric structure of an intact Paleozoic collisional orogen in the Ural Mountains. Hybrid-source seismic reflection and refraction data provide images of a crustal-scale collisional fabric and a pronounced crustal root preserved since Paleozoic time. Mantle reflections are observed at depths of more than 150 kilometers, possibly representing the base of
Uploads