Image annotation using aggregated page information from active and inactive indices
View Patent ↗Architecture that addresses page information lost as part of a selection process in a search engine framework. An aggregation process collects all page or document information from the same image cluster and uses the aggregated page information to annotate one or more selected image-page pairs within the same image cluster. Once the entire set of descriptive terms is received, the entire set of descriptive terms or only an optimum set of top N descriptive terms of the entire set is for annotation of one or more of the representative images in the cluster.
1. A system, comprising:
an aggregation component configured to aggregate all page information of an image cluster into aggregated page information, the image cluster created based on image similarity of images in the image cluster;
a selection component configured to select descriptive terms from the aggregated page information to represent the image cluster, the selection component selects same or different sets of descriptive terms for different image-page tuples of the image cluster;
an annotation component configured to annotate selected image-page tuples of the image cluster with the descriptive terms; and
at least one hardware processor configured to execute computer-executable instructions in a memory associated with the aggregation component, the selection component, and the annotation component.
2. The system of claim 1 , further comprising an indexing component configured to index the selected image-page tuples annotated with one or more of the descriptive terms.
3. The system of claim 1 , further comprising a training component configured to train a model that is employed to assign weights to the descriptive terms of pages associated with the page information.
4. The system of claim 3 , wherein the model is a statistical model configured to assign different weights to terms obtained from different locations of the pages.
5. The system of claim 3 , wherein the training component is configured to resolve term duplication weighting and cross-cluster term weighting.
6. The system of claim 1 , wherein the selection component is configured to select top weighted descriptive terms of the aggregated page information as the descriptive terms.
7. The system of claim 1 , wherein the selection component comprises a feature selection algorithm configured to select a set of terms having highest scores.
8. The system of claim 1 , further comprising a testing component configured to compute an optimum system operating state that is a compromise between system performance and ranking relevance performance, the optimum operating state obtained by selection of an optimum operating set of the descriptive terms.
9. A method, comprising acts of:
aggregating all page information of image-page tuples of an image cluster into aggregated page information;
selecting descriptive terms from the aggregated page information to represent the image cluster; and
annotating selected image-page tuples of the image cluster with the descriptive terms.
10. The method of claim 9 , further comprising indexing the selected image-page tuples.
11. The method of claim 9 , further comprising training a model that is employed to assign weights to the descriptive terms of the pages.
12. The method of claim 9 , further comprising selecting top weighted descriptive terms of the aggregated page information as the descriptive terms.
13. The method of claim 9 , further comprising selecting an optimum set of the descriptive terms based on system performance tradeoffs.
14. The method of claim 9 , further comprising selecting a set of terms having highest scores using a feature selection algorithm.
15. The method of claim 9 , further comprising resolving term duplication and cross-cluster term weighting issues.
16. A computer-readable storage medium comprising computer-executable instructions that when executed by a hardware processor, cause the processor to perform acts of:
aggregating all page information image-page tuples of an image cluster into aggregated page information;
selecting descriptive terms from the aggregated page information to represent the image cluster;
annotating selected image-page tuples of the image cluster with the descriptive terms; and
indexing the selected image-page tuples based on the descriptive terms.
17. The computer-readable storage medium of claim 16 , further comprising training a statistical model that assigns different term weights based on location of the descriptive terms in a page.
18. The computer-readable storage medium of claim 16 , further comprising selecting top weighted descriptive terms of the aggregated page information as the descriptive terms.
19. The computer-readable storage medium of claim 16 , further comprising selecting an optimum set from the top weighted descriptive terms based on system performance tradeoffs.
20. The computer-readable storage medium of claim 16 , further comprising computing an optimum system operating state based on derivation of an optimum set of the descriptive terms.