METHOD OF IDENTIFYING AN OBJECT IN A VISUAL SCENE
First Claim
1. A method of identifying an object in a visual scene, comprising:
- a. extracting at least a portion of an image containing information about an object;
b. determining a first plurality of features of said at least said portion of said image;
c. processing said first plurality of features with an inclusive neural network, wherein said inclusive neural network is adapted to provide a second plurality of inclusive probability values responsive to said first plurality of features, and a third plurality of at least two of said second plurality of inclusive probability values represent a probability that said at least said portion of said image corresponds to a corresponding at least one of at least two different classes of objects;
d. processing said first plurality of features with a fourth plurality of exclusive neural networks, wherein said fourth plurality is equal in number to said third plurality, each exclusive neural network of said fourth plurality of exclusive neural networks provides a corresponding first exclusive probability value representing a probability that said at least said portion of said image corresponds to one of said at least two different classes of objects, each said exclusive neural network provides a corresponding second exclusive probability value representing a probability that said at least said portion of said image does not correspond to said one of said at least two different classes of objects, and different exclusive neural networks of said fourth plurality of exclusive neural networks provide said corresponding first and second probabilities for corresponding different classes of objects of said at least two different classes of objects; and
e. identifying whether said at least said portion of said image corresponds to any of said at least two different classes of objects, or whether said at least said portion of said image does not correspond to said any of said at least two different classes of objects, responsive to said second plurality of inclusive probability values from said inclusive neural network, and responsive to said corresponding first and second exclusive probability values from each of said fourth plurality of exclusive neural networks.
4 Assignments
0 Petitions
Accused Products
Abstract
A plurality of features determined from at least a portion of an image containing information about an object are processed with an inclusive neural network, and with a plurality of exclusive neural networks, so as to provide a plurality of inclusive probability values representing probabilities that the portion of the image corresponds to at least one of at least two different classes of objects, and for each exclusive neural network, so as to provide first and second exclusive probability values representing probabilities that the portion of the image respectively corresponds. or not. to at least one class of objects. The plurality of inclusive probability values, and the first and second exclusive probability values from each of the exclusive neural networks, provide for identifying whether the portion of the image corresponds, or not, to any of the at least two different classes of objects.
-
Citations
35 Claims
-
1. A method of identifying an object in a visual scene, comprising:
-
a. extracting at least a portion of an image containing information about an object; b. determining a first plurality of features of said at least said portion of said image; c. processing said first plurality of features with an inclusive neural network, wherein said inclusive neural network is adapted to provide a second plurality of inclusive probability values responsive to said first plurality of features, and a third plurality of at least two of said second plurality of inclusive probability values represent a probability that said at least said portion of said image corresponds to a corresponding at least one of at least two different classes of objects; d. processing said first plurality of features with a fourth plurality of exclusive neural networks, wherein said fourth plurality is equal in number to said third plurality, each exclusive neural network of said fourth plurality of exclusive neural networks provides a corresponding first exclusive probability value representing a probability that said at least said portion of said image corresponds to one of said at least two different classes of objects, each said exclusive neural network provides a corresponding second exclusive probability value representing a probability that said at least said portion of said image does not correspond to said one of said at least two different classes of objects, and different exclusive neural networks of said fourth plurality of exclusive neural networks provide said corresponding first and second probabilities for corresponding different classes of objects of said at least two different classes of objects; and e. identifying whether said at least said portion of said image corresponds to any of said at least two different classes of objects, or whether said at least said portion of said image does not correspond to said any of said at least two different classes of objects, responsive to said second plurality of inclusive probability values from said inclusive neural network, and responsive to said corresponding first and second exclusive probability values from each of said fourth plurality of exclusive neural networks. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35)
-
Specification