Artificial Neural Network Incorporating Emphasis And Focus Techniques
First Claim
1. A method of implementing a convolutional neural network (CNN), the method comprising:
- receiving a frame buffer of pixels representing an input image;
convolving a first convolutional kernel with said frame buffer of pixels in accordance with a spatial aware function; and
wherein said spatial aware function is adapted to emphasize and focus on one or more portions of said input image whereby portions of said image having higher spatial significance are convolved before portions having lower spatial significance.
1 Assignment
0 Petitions
Accused Products
Abstract
A novel and useful artificial neural network that incorporates emphasis and focus techniques to extract more information from one or more portions of an input image compared to the rest of the image. The ANN recognizes that valuable information in an input image is typically not distributed throughout the image but rather is concentrated in one or more regions. Rather than implement CNN layers sequentially (i.e. row by row) on the input domain of each layer, the present invention leverages the fact that valuable information is focused in one or more regions of the image where it is desirable to apply more attention and for which it is desired to apply more elaborate evaluation. Precision dilution can be applied to those portions of the input image that are not the center of focus and emphasis. A spatial aware function determines the location(s) of the ears of focus and is applied to the first convolutional layer. Dilution of precision is performed either before and/or after the first convolutional layer thereby significantly reducing computation and power requirements.
5 Citations
20 Claims
-
1. A method of implementing a convolutional neural network (CNN), the method comprising:
-
receiving a frame buffer of pixels representing an input image; convolving a first convolutional kernel with said frame buffer of pixels in accordance with a spatial aware function; and wherein said spatial aware function is adapted to emphasize and focus on one or more portions of said input image whereby portions of said image having higher spatial significance are convolved before portions having lower spatial significance. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11)
-
-
12. A convolutional neural network (CNN) having emphasis and focus capability, comprising:
-
a frame buffer of pixels representing an input image; a convolution circuit operative to convolve a first convolutional kernel with said frame buffer of pixels in accordance with a spatial aware function; and wherein said spatial aware function is adapted to emphasize and focus on one or more portions of said input image whereby portions of said image having higher spatial significance are convolved before portions having lower spatial significance. - View Dependent Claims (13, 14, 15, 16, 17, 18, 19)
-
-
20. A method of implementing an artificial neural network (ANN), the method comprising:
-
receiving an input image comprising a plurality of pixels; reordering a sequence of operations on said input image such that pixels of higher spatial importance are operated on first; and applying precision dilution to said input image after and/or before said operation thereby extracting more information from those portions of emphasis and focus compared with the rest of the input image.
-
Specification