Computerized method and system for automated determination of high quality digital content
Disclosed are systems and methods for improving interactions with and between computers in a content generating, hosting and/or providing system supported by or configured with personal computing devices, servers and/or platforms. The systems interact to identify and retrieve data across platforms, which can be used to improve the quality of data used in processing interactions between or among processors in such systems. The disclosed systems and methods provide systems and methods for automatic discovery of high quality digital content. According to embodiments, the present disclosure describes improved computer system and methods directed to analyzing raw image data, such as features and descriptors of images in order to identify a high quality image(s). Such images can be identified from a database of images, and such images can be identified in real-time, or near real-time during the capture of an image(s) by a camera.
1. A method comprising:
analyzing, via a computing device, a collection of user generated content (UGC) images to identify a first image, said first image having associated social data indicating user interest in the first image below a social threshold;
parsing, via the computing device, the first image to extract raw image data, said raw image data comprising features associated with content of the first image;
identifying, via the computing device, a set of second images from the UGC collection, said second set of images being high-quality images, said identifying comprising identifying the raw image data of each of the second images;
comparing, via the computing device, the raw image data of the first image with the raw image data of the second images, said comparison comprising identifying a similarity between the raw image data of the first image and the raw image data of each second image in accordance with a comparison threshold;
determining, via the computing device, whether the first image is a high-quality image based on the comparison, said high-quality determination based on whether the similarity between the raw image data of the first image and the raw image data of each second image satisfies the comparison threshold; and
communicating, via the computing device, information associated with the first image to at least one user when said first image is determined to be a high-quality image.
2. The method of claim 1 , further comprising:
monitoring, over a network, said social data of the first image when said determination indicates the first image is not a high-quality image;
based on said monitoring, comparing the social data to said social threshold, said comparison comprising determining whether the social data has increased to satisfy the social threshold during said monitoring; and
communicating information associated with the first image to said at least one user when said social data of the first image has increased to satisfy the social threshold.
3. The method of claim 1 , wherein said high-quality determination comprises computing a score for the first image based on said comparison, said score indicating whether the first image is determined to be a high-quality image, wherein said score is associated with said first image in said UGC collection.
4. The method of claim 1 , wherein said comparison further comprises:
extracting the features of the first image based on the raw image data of the first image;
extracting features of each of the second images based on the second images' raw image data;
translating the features for the first image into a feature vector having a dimensional value proportional to a number of features;
translating the features, for each second image, into a feature vector having a dimensional value proportional to a number of features of each second image; and
determining said similarity based on a semantic similarity between each data point of the feature vector of the first image and each data point of the second feature vectors.
5. The method of claim 1 , further comprising:
determining a content category of the first image based on the first raw image data, wherein said second set of images corresponds to the determined content category.
6. The method of claim 1 , wherein said identifying the second set of images comprises:
identifying a set of third images from the UGC collection based on social data associated with each third image, said social data of each third image satisfying the social threshold thereby indicating a level of interest from users on the network above a quality threshold;
parsing each third image to extract raw image data for each third image, said raw image data for each third image comprising features associated with content of the respective third image;
analyzing a fourth set of images from the UGC collection based on the raw image data of the third set of images, said analysis comprising comparing the raw image data of the third set of images to raw image data of the fourth set of images; and
identifying said second set of images as a subset of images from the fourth set of images based on said comparison, said raw image data of the second set of images matching the raw image data of the third set of images at or above the quality threshold.
7. The method of claim 6 , wherein said comparison comprises:
determining a probability that the raw image data of the fourth set of images corresponds to the raw image data of the third set of images, said probability determination resulting in a score; and
comparing said score to the threshold, wherein satisfaction of the threshold indicates that the second set of images are high quality.
8. The method of claim 6 , further comprising:
parsing each fourth image to extract the raw image data for each fourth image, said raw image data for each fourth image comprising features associated with content of the respective fourth image, wherein said analysis further comprises formulating feature vectors based on the image data of the third and fourth image set, wherein said comparison is based on a comparison between each third feature vector and each fourth feature vector.
9. The method of claim 6 , wherein said indicated level of interest above the quality threshold provides an indication that the third set of images are high-quality images.
10. The method of claim 1 , further comprising:
determining a context of the first image based on the raw image data;
causing communication, over the network, of said context to an advertisement platform to obtain an advertisement associated with said context; and
communicating said identified advertisement in association with said communication of said first image.
11. A non-transitory computer-readable storage medium tangibly encoded with computer-executable instructions, that when executed by a processor associated with a computing device, performs a method comprising:
analyzing a collection of user generated content (UGC) images to identify a first image, said first image having associated social data indicating user interest in the first image below a social threshold;
parsing the first image to extract raw image data, said raw image data comprising features associated with content of the first image;
identifying a set of second images from the UGC collection, said second set of images being high-quality images, said identifying comprising identifying the raw image data of each of the second images;
comparing the raw image data of the first image with the raw image data of the second images, said comparison comprising identifying a similarity between the raw image data of the first image and the raw image data of each second image in accordance with a comparison threshold;
determining whether the first image is a high-quality image based on the comparison, said high-quality determination based on whether the similarity between the raw image data of the first image and the raw image data of each second image satisfies the comparison threshold; and
communicating information associated with the first image to at least one user when said first image is determined to be a high-quality image.
12. A system comprising:
a processor;
a non-transitory computer-readable storage medium for tangibly storing thereon program logic for execution by the processor, the program logic comprising:
analysis logic executed by the processor for analyzing a collection of user generated content (UGC) images to identify a first image, said first image having associated social data indicating user interest in the first image below a social threshold;
extraction logic executed by the processor for parsing the first image to extract raw image data, said raw image data comprising features associated with content of the first image;
identification logic executed by the processor for identifying a set of second images from the UGC collection, said second set of images being high-quality images, said identifying comprising identifying the raw image data of each of the second images;
comparison logic executed by the processor for comparing the raw image data of the first image with the raw image data of the second images, said comparison comprising identifying a similarity between the raw image data of the first image and the raw image data of each second image in accordance with a comparison threshold;
determination logic executed by the processor for determining whether the first image is a high-quality image based on the comparison, said high-quality determination based on whether the similarity between the raw image data of the first image and the raw image data of each second image satisfies the comparison threshold; and
communication logic executed by the processor for communicating information associated with the first image to at least one user when said first image is determined to be a high-quality image.