- Number
- 10469905
- Published
- 2019-11-05
- Filed
- 2018-08-03
- Assignee
- Disney Enterprises, Inc.
- Inventors
- Farre Guiu; Miquel Angel et al.
- CPC
- H04N21/23418; H04N21/4516; H04N21/44008; H04N21/435; G06V20/41; H04N21/2353
- Verdict
- Set aside video asset classification, content management
- Source
- Google Patents · FreePatentsOnline
Abstract
According to one implementation, a content classification system includes a computing platform having a hardware processor and a system memory storing a video asset classification software code. The hardware processor executes the video asset classification software code to receive video clips depicting video assets and each including images and annotation metadata, and to preliminarily classify the images with one or more of the video assets to produce image clusters. The hardware processor further executes the video asset classification software code to identify key features data corresponding respectively to each image cluster, to segregate the image clusters into image super-clusters based on the key feature data, and to uniquely identify each of at least some of the image super-clusters with one of the video assets.
Background
BACKGROUND(1) Video has for some time been, and continues to be a highly popular medium for the enjoyment of entertainment content in the form of movie, television, and sports content, for example, as well as for information content such as news. Due to its popularity with consumers, ever more video content is being produced and made available for distribution. Consequently, the accuracy and efficiency with which video content can be reviewed, classified, archived, and managed has become increasingly important to producers, owners, and distributors of such content. For example, techniques for automating the classification of video content based on features or images included in the video, may reduce the time spent in video production and management.(2) Unfortunately, conventional approaches to automating video classification typically require initial datasets that may be costly and time consuming to prepare. For example, conventional approaches to classifying video content based on image recognition require that collections of precisely labeled images be prepared as an initial input for comparative purposes.SUMMARY(3) There are provided video asset classification systems and methods, substantially as shown in and/or described in connection with at least one of the figures, and as set forth more completely in the claims.
Claims
1. A content classification system comprising: a computing platform including a hardware processor and a system memory; a video asset classification software code stored in the system memory; the hardware processor configured to execute the video asset classification software code to: produce a plurality of image clusters by classifying a plurality of images included in a plurality of video clips with at least one of a plurality of video assets known to be included in the plurality of video clips; identify key features data corresponding respectively to each image cluster of the plurality of image clusters; segregate the plurality of image clusters into image super-clusters based on the key features data, each of the image super-clusters including one or more of the plurality of image clusters; identify each of at least some of the image super-clusters with one of the plurality of video assets known to be included in the plurality of video clips; and train the video asset classification software code using the identified image super-clusters.
2. The content classification system of claim 1, wherein identifying each of the at least one of the image super-clusters with one of the plurality of video assets is based on a confidence value associated with classifications produced by the classifying of the plurality of images.
3. The content classification system of claim 1, wherein the plurality of video assets depicted in the video clips comprise previously identified video assets.
4. The content classification system of claim 1, wherein at least some of the first plurality of video clips comprise multiple shots.
5. The content classification system of claim 1, wherein more than one of the image clusters is classified with a same one of the plurality of video assets depicted in the video clips.
6. The content classification system of claim 1, wherein the plurality of video assets comprise dramatic characters.
7. The content classification system of claim 1, wherein the plurality of video assets comprise at least one of objects and locations.
8. A method for use by a content classification system including a computing platform having a hardware processor and a system memory storing a video asset classification software code for execution by the hardware processor, the method comprising: producing, using the hardware processor, a plurality of image clusters by classifying a plurality of images included in a plurality of video clips with at least one of a plurality of video assets known to be included in the plurality of video clips; identifying, using the hardware processor, key features data corresponding respectively to each image cluster of the plurality of image clusters; segregating, using the hardware processor, the plurality of image clusters into image super-clusters based on the key features data, each of the image super-clusters including one or more of the plurality of image clusters; identifying, using the hardware processor, each of at least some of the image super-clusters with one of the plurality of video assets known to be included in the plurality of video clips; and training, using the hardware processor, the video asset classification software code using the identified image super-clusters.
9. The method of claim 8, wherein identifying at least one of the image super-clusters with one of the plurality of video assets is based on a confidence value associated with classifications produced by the classifying of the plurality of images.
10. The method of claim 8, wherein the plurality of video assets depicted in the video clips comprise previously identified video assets.
11. The method of claim 8, wherein at least some of the first plurality of video clips comprise multiple shots.
12. The method of claim 8, wherein more than one of the image clusters is classified with a same one of the plurality of video assets depicted in the video clips.
13. The method of claim 8, wherein the plurality of video assets comprise dramatic characters.
14. The method of claim 8, wherein the plurality of video assets comprise at least one of objects and locations.
15. A computer-readable non-transitory medium having stored thereon instructions, which when executed by a hardware processor, instantiate a method comprising: producing a plurality of image clusters by classifying a plurality of images included in a plurality of video clips with at least one of a plurality of video assets known to be included in the plurality of video clips; identifying key features data corresponding respectively to each image cluster of the plurality of image clusters; segregating the plurality of image clusters into image super-clusters based on the key features data, each of the image super-clusters including one or more of the plurality of image clusters; identifying each of at least some of the image super-clusters with one of the plurality of video assets known to be included in the plurality of video clips; and training the video asset classification software code using the identified image super-clusters.
16. The computer-readable non-transitory medium of claim 15, wherein identifying at least one of the image super-clusters with one of the plurality of video assets is based on a confidence value associated with classifications produced by the classifying of the plurality of images.
17. The computer-readable non-transitory medium of claim 15, wherein the plurality of video assets depicted in the video clips comprise previously identified video assets.
18. The computer-readable non-transitory medium of claim 15, wherein at least some of the first plurality of video clips comprise multiple shots.
19. The computer-readable non-transitory medium of claim 15, wherein more than one of the image clusters is classified with a same one of the plurality of video assets depicted in the video clips.
20. The computer-readable non-transitory medium of claim 15, wherein the plurality of video assets comprise at least one of dramatic characters, objects, and locations.