Identification of complex shapes using a self organizing neural system

2000 ◽  
Vol 11 (4) ◽  
pp. 921-934 ◽  
Author(s):  
T. Sabisch ◽  
A. Ferguson ◽  
H. Bolouri
Keyword(s):  
Neural System ◽  
Complex Shapes ◽  
Self Organizing

scholarly journals Brain-Inspired Self-Organization with Cellular Neuromorphic Computing for Multimodal Unsupervised Learning

Electronics ◽  
2020 ◽  
Vol 9 (10) ◽  
pp. 1605 ◽  
Author(s):  
Lyes Khacef ◽  
Laurent Rodriguez ◽  
Benoît Miramond
Keyword(s):  
Unsupervised Learning ◽  
Cortical Plasticity ◽  
Internal Representation ◽  
Neural System ◽  
Self Organization ◽  
Self Organizing Map ◽  
Self Organizing Maps ◽  
Dynamic Vision ◽  
Neuromorphic Architecture ◽  
Self Organizing

Cortical plasticity is one of the main features that enable our ability to learn and adapt in our environment. Indeed, the cerebral cortex self-organizes itself through structural and synaptic plasticity mechanisms that are very likely at the basis of an extremely interesting characteristic of the human brain development: the multimodal association. In spite of the diversity of the sensory modalities, like sight, sound and touch, the brain arrives at the same concepts (convergence). Moreover, biological observations show that one modality can activate the internal representation of another modality when both are correlated (divergence). In this work, we propose the Reentrant Self-Organizing Map (ReSOM), a brain-inspired neural system based on the reentry theory using Self-Organizing Maps and Hebbian-like learning. We propose and compare different computational methods for unsupervised learning and inference, then quantify the gain of the ReSOM in a multimodal classification task. The divergence mechanism is used to label one modality based on the other, while the convergence mechanism is used to improve the overall accuracy of the system. We perform our experiments on a constructed written/spoken digits database and a Dynamic Vision Sensor (DVS)/EletroMyoGraphy (EMG) hand gestures database. The proposed model is implemented on a cellular neuromorphic architecture that enables distributed computing with local connectivity. We show the gain of the so-called hardware plasticity induced by the ReSOM, where the system’s topology is not fixed by the user but learned along the system’s experience through self-organization.

scholarly journals A self-organizing neural system for learning to recognize textured scenes

1999 ◽  
Vol 39 (7) ◽  
pp. 1385-1406 ◽  
Author(s):  
Stephen Grossberg ◽  
James R. Williamson
Keyword(s):  
Neural System ◽  
Self Organizing

Transmission Electron Diffraction Analysis by Computer

1970 ◽  
Vol 28 ◽  
pp. 40-41
Author(s):  
R.A. Ploc ◽  
G.H. Keech
Keyword(s):  
Electron Diffraction ◽  
Input Data ◽  
Reciprocal Lattice ◽  
Computer Programme ◽  
Transmission Electron Diffraction ◽  
Usual Procedure ◽  
Transmission Electron ◽  
Complex Shapes ◽  
Computer Input ◽  
Diffraction Effects

An unambiguous analysis of transmission electron diffraction effects requires two samplings of the reciprocal lattice (RL). However, extracting definitive information from the patterns is difficult even for a general orthorhombic case. The usual procedure has been to deduce the approximate variables controlling the formation of the patterns from qualitative observations. Our present purpose is to illustrate two applications of a computer programme written for the analysis of transmission, selected area diffraction (SAD) patterns; the studies of RL spot shapes and epitaxy.When a specimen contains fine structure the RL spots become complex shapes with extensions in one or more directions. If the number and directions of these extensions can be estimated from an SAD pattern the exact spot shape can be determined by a series of refinements of the computer input data.

The crystallization of thin-film ceramics

1992 ◽  
Vol 50 (1) ◽  
pp. 338-339
Author(s):  
J.M. Schwartz ◽  
L.F. Francis ◽  
L.D. Schmidt ◽  
P.S. Schabes-Retchkiman
Keyword(s):  
Thin Films ◽  
Grain Size ◽  
Sol Gel ◽  
Processing Route ◽  
Small Areas ◽  
Ceramic Thin Films ◽  
Complex Shapes ◽  
Sol Gel Process ◽  
Ceramic Precursors ◽  
Crystalline Ceramic

Ceramic thin films and coatings are of interest for electrical, optical, magnetic and thermal barrier applications. Critical for improved properties in thin films is the development of specific microstructures during processing. To this end, the sol-gel method is advantageous as a versatile processing route. The sol-gel process involves depositing a solution containing metalorganic or colloidal ceramic precursors onto a substrate and heating the deposited layer to form a crystalline or non-crystalline ceramic coating. This route has several advantages, including the ability to create tailored microstructures and properties, to coat large or small areas, simple or complex shapes, and to more easily prepare multicomponent ceramics. Sol-gel derived coatings are amorphous in the as-deposited state and develop their crystalline structure and microstructure during heat-treatment. We are particularly interested in studying the amorphous to crystalline transformation, because many key features of the microstructure such as grain size and grain size distribution may be linked to this transformation.

Predicting spatial activity patterns in a neural map from a probabilistic atlas of 3-D reconstructed neurons

1995 ◽  
Vol 53 ◽  
pp. 812-813
Author(s):  
G. Jacobs ◽  
F. Theunissen
Keyword(s):  
Activity Patterns ◽  
Three Dimensional ◽  
Sensory System ◽  
Neural System ◽  
Probabilistic Atlas ◽  
Uniform Sample ◽  
Dimensional Reconstruction ◽  
The Time Domain ◽  
And Function ◽  
Visualization Techniques

In order to understand how the algorithms underlying neural computation are implemented within any neural system, it is necessary to understand details of the anatomy, physiology and global organization of the neurons from which the system is constructed. Information is represented in neural systems by patterns of activity that vary in both their spatial extent and in the time domain. One of the great challenges to microscopists is to devise methods for imaging these patterns of activity and to correlate them with the underlying neuroanatomy and physiology. We have addressed this problem by using a combination of three dimensional reconstruction techniques, quantitative analysis and computer visualization techniques to build a probabilistic atlas of a neural map in an insect sensory system. The principal goal of this study was to derive a quantitative representation of the map, based on a uniform sample of afferents that was of sufficient size to allow statistically meaningful analyses of the relationships between structure and function.

Sign in / Sign up

Export Citation Format

Share Document