Image analysis – Learning systems – Neural networks
Reexamination Certificate
2001-05-10
2004-10-05
Mehta, Bhavesh M. (Department: 2625)
Image analysis
Learning systems
Neural networks
C382S305000, C701S059000, C704S232000, C706S015000, C706S042000
Reexamination Certificate
active
06801655
ABSTRACT:
BACKGROUND OF THE INVENTION
(1) Field of the Invention
The invention relates to neural networks and is directed more particularly to a spatial image processor neural network for processing spatial image data to distinguish one configuration of component objects from a different configuration of the same component objects.
(2) Description of the Prior Art
Electronic neural networks have been developed to rapidly identify patterns in certain types of input data, or to classify accurately the input patterns into one of a plurality of predetermined classifications. For example, neural networks have been developed which can recognize and identify patterns, such as the identification of hand-written alphanumeric characters, in response to input data constituting the pattern of on/off picture elements, or “pixels,” representing the images of the characters to be identified. In such a neural network, the pixel pattern is represented by, for example, electrical signals coupled to a plurality of input terminals, which, in turn, are connected to a number of processing nodes, each of which is associated with one of the alphanumeric characters which the neural network can identify. The input signals from the input terminals are coupled to the processing nodes through certain weighting functions, and each processing node generates an output signal which represents a value that is a non-linear function of the pattern of weighted input signals applied thereto. Based on the values of the weighted pattern of input signals from the input terminals, if the input signals represent a character, which can be identified by the neural network, one of the processing nodes, which is associated with that character will generate a positive output signal, and the others will not. On the other hand, if the input signals do not represent a character, which can be identified by the neural network, none of the processing nodes will generate a positive output signal. Neural networks have been developed which can perform similar pattern recognition in a number of diverse areas.
The particular patterns that the neural network can identify depend on the weighting functions and the particular connections of the input terminals to the processing nodes. As an example, the weighting functions in the above-described character recognition neural network essentially will represent the pixel patterns that define each particular character. Typically, each processing node will perform a summation operation in connection with values representing the weighted input signals provided thereto, to generate a sum that represents the likelihood that the character to be identified is the character associated with that processing node. The processing node then applies the nonlinear function to that sum to generate a positive output signal if the sum is, for example, above a predetermined threshold value. Conventional nonlinear functions which processing nodes may use in connection with the sum of weighted input signals generally include a step function, a threshold function, or a sigmoid. In all cases the output signal from the processing node will approach the same positive output signal asymptotically.
Before a neural network can be useful, the weighting functions for each of the respective input signals must be established. In some cases, the weighting functions can be established a priori. Normally, however, a neural network goes through a training phase in which input signals representing a number of training patterns for the types of items to be classified (e.g., the pixel patterns of the various hand-written characters in the character-recognition example) are applied to the input terminals, and the output signals from the processing nodes are tested. Based on the pattern of output signals from the processing nodes for each training example, the weighting functions are adjusted over a number of trials. After being trained, the neural network can generally accurately recognize patterns during an operational phase, with the degree of success based in part on the number of training patterns applied to the neural network during the training stage, and the degree of dissimilarity between patterns to be identified. Such a neural network can also typically identify patterns that are similar, but not necessarily identical, to the training patterns.
One of the problems with conventional neural network architectures as described above is that the training methodology, generally known as the “back-propagation” method, is often extremely slow in a number of important applications. In addition, under the back-propagation method, the neural network may result in erroneous results, which may require restarting of training. Even after a neural network has been through a training phase confidence that the best training has been accomplished may sometimes be poor. If a new classification is to be added to a trained neural network, the complete neural network must be retrained. In addition, the weighting functions generated during the training phase often cannot be interpreted in ways that readily provide understanding of what they particularly represent.
Thus, a neural network is typically considered to be a trainable entity that can be taught to transform information for a purpose. Neural networks are adaptable through a form of training, which is usually by example. Long training times is a problem in trainable neural networks.
The spatial image processor is part of a new neural network technology that is constructed rather than trained as in common neural networks. Since the words “neural network” often connote a totally trainable neural network, the full definition of a constructed neural network, as used herein, is as follows: A constructed neural network is a connectionist neural network system that is assembled using common neural network components to perform a specific process. The assembly is analogous to the construction of an electronic assembly using resistors, transistors, integrated circuits and other simple electronic parts. Some examples of common neural components are specific values and/or types of connections, processing elements (neurons), output functions, gain elements and other artificial neural network parts. As in electronics, the laws of nature, such as mathematics, physics, chemistry, mechanics, and “Rules of Experience” govern the assembly and architecture of a constructed neural network. A constructed neural network, which is assembled for a specific process without the necessity of training, can be considered equivalent to a trained common neural network with an infinite training sequence that has attained an output error of zero. Most neural network systems of many constructed neural network modules, such as the spatial objects data fuser, have weights that are never altered after they are constructed. When the traditional neural network system is trained, learning occurs only in special memory modules. Such special memory modules are part of this new constructed neural network technology that learns an example in a single application and does not require a retraining of the old examples when a new example is added to a previously trained system, i.e., old memory is retained and not altered.
In artificial neural networks various neural components have synonyms. For example a “neuron”, a “processing element” and a “processing node” are the same. A “connection value”, a “weight value” and “weighting value” are the same. One or more of such synonyms are used in this and or other associated applications.
Despite advances in spatial image processors, there remains a need for a spatial image processor neural network wherein the spatial image processor neural network has a very high neuron count (approximately 10
5
to 10
8
neurons), depending on the multidimensional space the neural network modules operate, and is of an architectural structure providing unique attributes:
(1) The spatial image processor discriminates between two groups comprised of identical components in two different spatial configurations. It is n
Chawan Sheela
Kasischke James M.
Mehta Bhavesh M.
Nasser Jean-Paul A.
Oglo Michael F.
LandOfFree
Spatial image processor does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Spatial image processor, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Spatial image processor will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3312672