Reduction of search ambiguity with multiple media references
Methods and apparatus for implementing a technique for searching media objects. In general, in one aspect, the technique includes receiving user input specifying a plurality of reference objects, defining a set of features for them, and combining the features to generate composite reference information defining criteria for search. In general, in another aspect, the technique includes combining object information for a plurality of reference objects to produce composite reference information, comparing the composite reference information to object information for media objects in a collection of media objects, and selecting a media object based upon the comparison.
1. A computer-implemented method comprising:
receiving identification of two or more reference video objects;
calculating, by a computer, video parameters for each reference video object, wherein the video parameters characterize image features and audio features in each reference video object that are common to a first reference video object and at least a second reference video object;
combining, by the computer, the video parameters for the first reference video object and at least the second reference video object to generate composite reference information; and
comparing, by the computer, the composite reference information to video objects in a collection of video objects stored on one or more databases to identify one or more video objects in the collection having features described by the composite reference information.
2. The computer-implemented method of claim 1 , wherein the image features comprise texture, color, shape, or combinations thereof, and wherein the audio features comprise Fourier transform information, wavelet decomposition information, fractal measures, or combinations thereof.
3. The computer-implemented method of claim 1 , wherein the video parameters include fractal measures derived from an analysis of combined audio and visual aspects of the video object.
4. The computer-implemented method of claim 1 , wherein the video parameters are combined to encompass characteristics of any of the reference video objects.
5. The computer-implemented method of claim 1 , wherein the video parameters are combined to specify characteristics that are common to all of the reference video objects.
6. The computer-implemented method of claim 1 , further comprising receiving identification of a new reference object after identifying one or more video objects in the collection having the features described by the composite reference information.
7. The computer-implemented method of claim 6 , further comprising modifying the composite reference information based on the new reference object, wherein the new reference object is an audio object, an image object, a text object, or a video object.
8. The computer-implemented method of claim 1 , wherein the calculated video parameters are combined into a feature vector for each reference video object.
9. The computer-implemented method of claim 8 , wherein the feature vector includes multiple sets of image components and audio components.
10. The computer-implemented method of claim 8 , wherein the feature vector is standardized to sum to unity.
11. The computer-implemented method of claim 8 , wherein the feature vector is not standardized to sum to unity, and comparing comprises using a modified Min function to produce vectors for which a sum of components ranges from 0 to 1.
12. The computer-implemented method of claim 1 , wherein the identification of the two or more reference video objects is received from a user computing device to identify similar video objects within the collection of video objects, and wherein the method further comprises providing the one or more identified video objects from the collection of video objects for presentation at the user computing device.
13. A system comprising:
a processor configured to execute instructions;
a non-transitory memory; and
computer program instructions embodied on the memory, which, when executed by the processor, cause the processor to:
receive identification of two or more reference video objects;
calculate video parameters for each reference video object, wherein the video parameters characterize image features and audio features in each reference video object that are common to a first reference video object and at least a second reference video object;
combine the video parameters for the first reference video object and at least the second reference video object to generate composite reference information; and
compare the composite reference information to video objects in a collection of video objects stored on one or more databases to identify one or more video objects in the collection having features described by the composite reference information.
14. The system of claim 13 , wherein the audio features comprise Fourier transform information, wavelet decomposition information, fractal measures, or combinations thereof.
15. The system of claim 13 , wherein the video parameters include fractal measures derived from an analysis of combined audio and visual aspects of the video object.
16. The system of claim 13 , wherein the calculated video parameters are combined into a feature vector for each reference video object, and wherein the feature vector includes multiple sets of image components and audio components.
17. The system of claim 16 , wherein the feature vector is standardized to sum to unity, or, if the feature vector is not standardized to sum to unity, using a modified Min function to produce vectors for which a sum of components ranges from 0 to 1.
18. The system of claim 13 , wherein the identification of the two or more reference video objects is received from a user computing device to identify similar video objects within the collection of video objects, and wherein the processor is further caused to provide the one or more identified video objects from the collection of video objects for presentation at the user computing device.
19. Computer program instructions, encoded on one or more non-transitory computer-readable memories, which, when executed by a computer processor, cause the processor to:
receive identification of two or more reference video objects;
calculate video parameters for each reference video object, wherein the video parameters characterize image features and audio features in each reference video object that are common to a first reference video object and at least a second reference video object;
combine the video parameters for the first reference video object and at least the second reference video object to generate composite reference information; and
compare the composite reference information to video objects in a collection of video objects stored on one or more databases to identify one or more video objects in the collection having features described by the composite reference information.
20. The computer program instructions of claim 19 , wherein the identification of the two or more reference video objects is received from a user computing device to identify similar video objects within the collection of video objects, and wherein the processor is further caused to provide the one or more identified video objects from the collection of video objects for presentation at the user computing device.