Automated image ads

1. A system, comprising:

a database storing a plurality of text advertisements comprising a text advertisement; and

an advertisement server configured to:

access the plurality of text advertisements;

receive, via a first network communication, an advertisement request from a client device;

select the text advertisement, from the database storing the plurality of text advertisements, based on the advertisement request;

responsive to selecting the text advertisement, determine an advertiser web site associated with serving content related to the text advertisement selected from the database based on the advertisement request;

retrieve, via a second network communication, a plurality of images from the advertiser web site associated with serving content related to the text advertisement;

provide at least one of a title or description of the text advertisement to a deep learning engine;

provide autotags of one or more images of the plurality of images to the deep learning engine;

perform, via the deep learning engine, at least one of text similarity or semantics similarity to determine relevance features;

rank the plurality of images based upon features, comprising the relevance features, determined using text of the text advertisement;

select an image, from the plurality of images retrieved from the advertiser web site, based on the ranking;

responsive to selecting the image from the plurality of images retrieved from the advertiser web site:

generate an enhanced advertisement, in response to the advertisement request from the client device, by integrating:

(i) the image selected from the plurality of images retrieved from the advertiser web site; with

(ii) the text advertisement selected from the database based on the advertisement request received from the client device; and

transmit, via a third network communication, the enhanced advertisement generated in response to the advertisement request to the client device.

2. The system of claim 1 , wherein the ranking is based upon feedback of user behavior associated with one or more advertisement placements that included the image.

3. The system of claim 1 , wherein the ranking is based upon a product detection algorithm which identifies a product in the one or more images.

4. The system of claim 1 , wherein the ranking is based upon a non-photo detection algorithm which determines if the image is a drawing, a sketch, text, a logo, a map or an illustration.

5. The system of claim 1 , wherein the ranking is based upon an adult content detection algorithm.

6. The system of claim 1 , wherein the ranking comprises ranking at least one image lower in response to a text recognition algorithm determining the image includes more than a threshold amount of text.

7. The system of claim 1 , wherein the ranking comprises ranking at least one image lower in response to a logo detection algorithm determining that the image is a logo.

8. The system of claim 1 , wherein the ranking comprises ranking at least one image lower in response to a face detection algorithm that determines when a face is a main feature in the image.

9. The system of claim 1 , wherein the ranking comprises ranking at least one image lower in response to an ad aesthetic algorithm that identifies a most aesthetically pleasing image based on feedback loop training on images that have been successful in increasing click rates.

10. The system of claim 1 , wherein the ranking comprises ranking at least one image lower in response to a smart cropping algorithm that sizes the image.

11. The system of claim 1 , wherein the ranking is based upon feedback of user behavior associated with the image.

12. The system of claim 11 , wherein the ranking is based upon feedback of clicks associated with the image.

13. The system of claim 11 , wherein the ranking is based upon feedback of conversions associated with the image.

14. The system of claim 1 , wherein the ranking is based upon device configuration information associated with image.

15. The system of claim 1 , wherein the ranking is based upon geo-location information associated with the image.

16. A method, comprising:

storing a plurality of text advertisements, comprising a text advertisement, in a database;

receiving, via a first network communication, an advertisement request from a client device;

selecting the text advertisement, from the database storing the plurality of text advertisements, based on the advertisement request;

responsive to selecting the text advertisement, determining an advertiser web site associated with serving content related to the text advertisement selected from the database based on the advertisement request;

retrieving, via a second network communication, a plurality of images from the advertiser web site associated with serving content related to the text advertisement;

providing text of the text advertisement to a deep learning engine;

providing autotags of one or more images of the plurality of images to the deep learning engine;

performing, via the deep learning engine, at least one of text similarity or semantics similarity to determine relevance features;

ranking the plurality of images based upon features, comprising the relevance features, determined using the text of the text advertisement;

selecting an image, from the plurality of images retrieved from the advertiser web site, based on the ranking; and

responsive to selecting the image from the plurality of images retrieved from the advertiser web site:

generate an enhanced advertisement, in response to the advertisement request from the client device, by integrating:

(i) the image selected from the plurality of images retrieved from the advertiser web site; with

(ii) the text advertisement selected from the database based on the advertisement request received from the client device.

17. The method of claim 16 , wherein the ranking is based upon a non-photo detection algorithm which determines if the image is a drawing, a sketch, text, a logo, a map or an illustration.

18. The method of claim 16 , wherein the ranking is based upon feedback of user behavior associated with one or more advertisement placements that included the image.

19. A system, comprising:

a processor configured to:

receive an advertisement request from a client device;

select a text advertisement from a database based on the advertisement request;

responsive to selecting the text advertisement, determine an advertiser web site associated with serving content related to the text advertisement selected from the database based on the advertisement request;

retrieve a plurality of images from the advertiser web site associated with serving content related to the text advertisement;

provide at least one of a title or description of the text advertisement to a deep learning engine;

provide autotags of one or more images of the plurality of images to the deep learning engine;

perform, via the deep learning engine, at least one of text similarity or semantics similarity to determine relevance features;

rank the plurality of images based upon features, comprising the relevance features, determined using text of the text advertisement;

select an image, from the plurality of images retrieved from the advertiser web site, based on the ranking;

responsive to selecting the image from the plurality of images retrieved from the advertiser web site:

generate an enhanced advertisement, in response to the advertisement request from the client device, by integrating:

(i) the image selected from the plurality of images retrieved from the advertiser web site; with

(ii) the text advertisement selected from the database based on the advertisement request received from the client device; and

transmit the enhanced advertisement generated in response to the advertisement request to the client device.

20. The system according to claim 19 , wherein the ranking is based upon a product detection algorithm which identifies a product in the one or more images.