Automatic generation of an image having an attribute from a subject image
The technology disclosed herein enables automatic generation of an image having an attribute using a subject image as a basis for the generated image. In a particular embodiment, a method includes executing a generator and a discriminator. Until a criterion is satisfied, the method includes iteratively performing steps a-c. In step a, the method includes providing an input vector and an attribute to the generator. The generator outputs a generated image. In step b, the method includes providing the generated image to the discriminator. The discriminator outputs feedback that indicates differences between the generated image and a subject image. In step c, the method includes determining whether the differences satisfy the criterion. When the differences do not satisfy the criterion, the input vector comprises the feedback in the next iteration of steps a-c. When the differences satisfy the criterion, the method includes associating the generated image with the attribute.
1 . A method comprising:
executing a generator and a discriminator;
until a criterion is satisfied, iteratively performing steps a-c:
a. providing an input vector and an attribute to the generator, wherein the generator outputs a generated image;
b. providing the generated image to the discriminator, wherein the discriminator outputs feedback that indicates differences between the generated image and a subject image;
c. determining whether the differences satisfy the criterion, wherein, when the differences do not satisfy the criterion, the input vector comprises the feedback in the next iteration of steps a-c; and
when the differences satisfy the criterion, associating the generated image with the attribute.
2 . The method of claim 1 , wherein in a first iteration of steps a-c, the input vector comprises a random input vector.
3 . The method of claim 1 , wherein the subject image comprises an image of a user and the attribute comprises an indicator of an emotion.
4 . The method of claim 3 , wherein, when the difference satisfies the criterion, the generated image comprises an image of the user expressing the emotion.
5 . The method of claim 1 , comprising:
until the criterion is satisfied, iteratively performing steps d-f:
d. providing a second input vector and a second attribute to the generator, wherein the generator outputs a second generated image;
e. providing the second generated image to the discriminator, wherein the discriminator outputs second feedback that defines second differences between the second generated image and the subject image;
f. determining whether the second differences satisfy the criterion, wherein, when the second differences do not satisfy the criterion, the second input vector comprises the second feedback in the next iteration of steps a-c; and
when the second differences satisfy the criterion, associating the second generated image with the second attribute.
6 . The method of claim 1 , comprising:
transmitting the generated image over a communication session established between a first endpoint and a second endpoint.
7 . The method of claim 6 , wherein the subject image comprises an image of a user operating the first endpoint.
8 . The method of claim 7 , comprising:
before transmitting the generated image, receiving, from the user, a selection of the generated image from a plurality of generated images that are each associated with respective ones of a plurality of attributes.
9 . The method of claim 8 , comprising:
before receiving the selection, presenting the plurality of generated images to the user.
10 . The method of claim 7 , comprising:
before transmitting the generated image, identifying a context of user communications received from the user;
determining that the context corresponds to the attribute; and
selecting the generated image from a plurality of generated images based on the attribute.
11 . An apparatus for a contact center system, the apparatus comprising:
one or more computer readable storage media;
a processing system, including processing circuitry, operatively coupled with the one or more computer readable storage media; and
program instructions stored on the one or more computer readable storage media that, when read and executed by the processing system, direct the processing system to:
execute a generator and a discriminator;
until a criterion is satisfied, iteratively perform steps a-c:
a. provide an input vector and an attribute to the generator, wherein the generator outputs a generated image;
b. provide the generated image to the discriminator, wherein the discriminator outputs feedback that indicates differences between the generated image and a subject image;
c. determine whether the differences satisfy the criterion, wherein, when the differences do not satisfy the criterion, the input vector comprises the feedback in the next iteration of steps a-c; and
when the differences satisfy the criterion, associate the generated image with the attribute.
12 . The apparatus of claim 11 , wherein in a first iteration of steps a-c, the input vector comprises a random input vector.
13 . The apparatus of claim 11 , wherein the subject image comprises an image of a user and the attribute comprises an indicator of an emotion.
14 . The apparatus of claim 13 , wherein, when the differences satisfy the criterion, the generated image comprises an image of the user expressing the emotion.
15 . The apparatus of claim 11 , wherein the program instructions direct the processing system to:
until the criterion is satisfied, iteratively perform steps d-f:
d. provide a second input vector and a second attribute to the generator, wherein the generator outputs a second generated image;
e. provide the second generated image to the discriminator, wherein the discriminator outputs second feedback that defines second differences between the second generated image and the subject image;
f. determine whether the second differences satisfy the criterion, wherein, when the second differences do not satisfy the criterion, the second input vector comprises the second feedback in the next iteration of steps a-c; and
when the second differences satisfy the criterion, associate the second generated image with the second attribute.
16 . The apparatus of claim 11 , wherein the program instructions direct the processing system to:
transmit the generated image over a communication session established between a first endpoint and a second endpoint.
17 . The apparatus of claim 16 , wherein the subject image comprises an image of a user operating the first endpoint.
18 . The apparatus of claim 17 , wherein the program instructions direct the processing system to:
before the generated image is transmitted, receive, from the user, a selection of the generated image from a plurality of generated images that are each associated with respective ones of a plurality of attributes.
19 . The apparatus of claim 17 , wherein the program instructions direct the processing system to:
before the generated image is transmitted, identify a context of user communications received from the user;
determine that the context corresponds to the attribute; and
select the generated image from a plurality of generated images based on the attribute.
20 . One or more computer readable storage media having program instructions stored thereon that, when read and executed by a processing system of a contact center system, direct the contact center system to:
execute a generator and a discriminator;
for each of a plurality of attributes:
until a criterion is satisfied, iteratively perform steps a-c:
a. provide an input vector and an attribute of the plurality of attributes to the generator, wherein the generator outputs a generated image;
b. provide the generated image to the discriminator, wherein the discriminator outputs feedback that indicates differences between the generated image and a subject image;
c. determine whether the differences satisfy the criterion, wherein, when the differences do not satisfy the criterion, the input vector comprises the feedback in the next iteration of steps a-c; and
when the differences satisfy the criterion, associate the generated image with the attribute.