Advanced techniques for the prediction originating from conventional neural network approaches, the local risk minimizer introduced by Vapkin, showed that it consistently outperforms other conventional neural network ...
详细信息
Advanced techniques for the prediction originating from conventional neural network approaches, the local risk minimizer introduced by Vapkin, showed that it consistently outperforms other conventional neural network techniques. An ESPRIT project is presented to take advantage of the techniques described to perform efficient prediction in demanding finance applications. The conventional multi-layer perception and Vapkin's theory of local estimation are compared for the applications of prediction: one in a check processing center and the other, in a Credit Card Call Center, both operating at sites run by SLIGOS.
In this paper a neural-network technique for classification of blocks of discrete cosine transform (DCT) coefficients using a backpropagation algorithm is described. The DCT is employed in a variety of transform-based...
详细信息
In this paper a neural-network technique for classification of blocks of discrete cosine transform (DCT) coefficients using a backpropagation algorithm is described. The DCT is employed in a variety of transform-based image compression schemes. In the authors' recent JPEG-like image compression scheme, efficient reordering of coefficients is achieved by applying adaptive zigzag reordering to variable-size rectangular sub-blocks. The additional neural-network-based sub-block classification discards isolated nonzero coefficients of small significance in some sub-blocks and therefore further reduces their sizes. Initial experimental results are presented that demonstrate the potential of the additional neural-network-based sub-block classification in terms of improved coding gain.
We describe a general method for extraction of an image object based on an approximate knowledge of its shape and its position in the image. The method is combination of the watershed approach and neuralnetworks. It ...
详细信息
ISBN:
(纸本)085296692X
We describe a general method for extraction of an image object based on an approximate knowledge of its shape and its position in the image. The method is combination of the watershed approach and neuralnetworks. It is applied to high field MR images of a section of a spinal cord for the extraction of the shape of the grey matter.
In this paper a method that facilitates an iconic query of an image/video database is presented. A query object is characterised by colour and texture properties. A feed-forward neural network is then trained on these...
详细信息
ISBN:
(纸本)085296692X
In this paper a method that facilitates an iconic query of an image/video database is presented. A query object is characterised by colour and texture properties. A feed-forward neural network is then trained on these features using the conjugate gradient-descent algorithm. The same characteristics are computed locally for the database images. The trained neural network is then used to test for the similarity between the iconically specified query and the database image descriptors. We show that by carefully selecting the set of descriptors we can significantly reduce the network size whilst not affecting the quality of results obtained.
The backpropagation neural network is applied to three gauging fringe analysis applications: classification of five spherical surfaces of differing radii;classification of five real objects with surfaces of different ...
详细信息
ISBN:
(纸本)085296692X
The backpropagation neural network is applied to three gauging fringe analysis applications: classification of five spherical surfaces of differing radii;classification of five real objects with surfaces of different radii of curvature;and identification of eggs according to their given commercial grades. Three methods for creating test and training vectors are used, of which the fast Fourier transform demonstrated drastic reduction in the network size, while applications with relatively noise-free data are indicated for mean/standard deviation driven networks.
In this paper, a technique employing artificial neuralnetworks for post-processing block coded images is presented. Visually important image features are extracted from the decompressed image and used as input to a f...
详细信息
In this paper, a technique employing artificial neuralnetworks for post-processing block coded images is presented. Visually important image features are extracted from the decompressed image and used as input to a feedforward neural network. The neural network learns to reconstruct the difference image between the original (uncompressed) and the decompressed image. Coding artifact reduction is achieved by adding the neuralnetworks output to the de-compressed image. Experimental results using the new technique for post-processing quadtree coded images are presented. It is shown the new technique can significantly improve the compressed image both in terms of peak signal to noise ratio (PSNR) and visual quality of the image.
This paper describes the architecture and the operation of a neural network based system for image interpretation. The system is based on the use of two models of associative neuralnetworks, ADAM and AURA for image a...
详细信息
This paper describes the architecture and the operation of a neural network based system for image interpretation. The system is based on the use of two models of associative neuralnetworks, ADAM and AURA for image and symbolic processing respectively. Employing characteristics of cellular automata theory and applying ideas from syntactic and structural pattern recognition, it uses a hierarchical approach to learn the structure of images. The hardware implementation of this system is based on the C-NNAP hardware platform.
The effectiveness of using region-based feature data as keys to an image database query search is presented. The classification engine required for this form of query has been implemented using a radial-basis function...
详细信息
ISBN:
(纸本)085296692X
The effectiveness of using region-based feature data as keys to an image database query search is presented. The classification engine required for this form of query has been implemented using a radial-basis function (RBF) network. It provides significant results using low volume training data provided in the form of feedback from the user on the progress of the query.
Advanced visual surveillance can benefit from taking a purposive approach to system design by using representations that are closely tailored to the surveillance tasks. A new scheme within the Bayesian belief network ...
详细信息
Advanced visual surveillance can benefit from taking a purposive approach to system design by using representations that are closely tailored to the surveillance tasks. A new scheme within the Bayesian belief network (BBN) framework is developed to provide the combination of constraints appropriate for incremental recovery in the face of uncertain and incomplete visual evidence. BBN techniques also provide a means of performing both bottom-up, data driven processing and top-down, expectation driven processing in the on-line computation. BBNs allow the computation of the most-probable-explanation of visual evidence under the expectations at all levels of abstraction in a vision system.
The papers submitted to the Sixth International Conference on imageprocessing and Its applications are presented. The issues considered include shape description and recognition;imageprocessingapplications;texture;...
详细信息
The papers submitted to the Sixth International Conference on imageprocessing and Its applications are presented. The issues considered include shape description and recognition;imageprocessingapplications;texture;image segmentation;neuralnetworks;colour;inspection and document processing;filtering and morphology;medical applications;transport, security and remote sensing.
暂无评论