Efficient calculation of a robust signature of a media unit

Patent metadata

Field	Value
Publication number	US-12067756-B2
Application number	US-202017594001-A
Country	US
Kind code	B2
Filing date	Mar 25, 2020
Priority date	Mar 31, 2019
Publication date	Aug 20, 2024
Grant date	Aug 20, 2024

Abstract

Systems, and method and computer readable media that store instructions for calculating signatures, utilizing signatures and the like, wherein for a low-power calculation of a signature, the method comprises: receiving or generating a media unit of multiple objects: processing the media unit by performing multiple iterations, determining a relevancy of the spanning elements of the iteration; completing the dimension expansion process by relevant spanning elements of the iteration and reducing a power consumption of irrelevant spanning; determining identifiers that are associated with significant portions of an output of the multiple iterations; and providing a signature that comprises the identifiers and represents the multiple objects.

First claim

We claim:

1. A method for low-power calculation of a signature, the method comprises:
receiving or generating a media unit of multiple objects;
processing the media unit by performing multiple iterations, wherein for each iteration of at least some of the multiple iterations comprises applying, by spanning elements of the iteration, a dimension expansion process that is followed by a merge operation; and
wherein the applying of the dimension expansion process of the iteration comprises:
determining a relevancy of the spanning elements of the iteration;
completing the dimension expansion process by relevant spanning elements of the iteration and reducing a power consumption of irrelevant spanning elements until, at least, a completion of the applying of the dimension expansion process;

determining identifiers that are associated with significant portions of an output of the multiple iterations; and
providing a signature that comprises the identifiers and represents the multiple objects.

2. The method according to claim 1 wherein the identifiers are retrieval information for retrieving the significant portions.

3. The method according to claim 1 wherein the at least some of the multiple iterations are a majority of the multiple iterations.

4. The method according to claim 1 wherein the output of the multiple iteration comprises multiple property attributes for each segment out of multiple segments of the media unit; and wherein the significant portions of an output of the multiple iterations comprises more impactful property attributes.

5. The method according to claim 1 wherein a first iteration of the multiple iteration comprises applying the dimension expansion process by applying different filters on the media unit.

6. The method according to claim 1 wherein the at least some of the multiple iteration exclude at least a first iteration of the multiple iterations.

7. The method according to claim 1 wherein the determining the relevancy of the spanning elements of the iteration is based on at least some identities of relevant spanning elements of at least one previous iteration.

8. The method according to claim 1 wherein the determining the relevancy of the spanning elements of the iteration is based on at least some identities of relevant spanning elements of at least one previous iteration that preceded the iteration.

9. The method according to claim 1 wherein the determining the relevancy of the spanning elements of the iteration is based on properties of the media unit.

10. The method according to claim 1 wherein the determining the relevancy of the spanning elements of the iteration is performed by the spanning elements of the iteration.

11. The method according to claim 1 comprising a neural network processing operation that is executed by one or more layers of a neural network and does not belong to the at least some of the multiple iterations.

12. The method according to claim 11 , wherein the at least one iteration is executed without reducing power consumption of irrelevant neurons of the one or more layers.

13. The method according to claim 11 comprising outputting, by the one or more layers, information about properties of the media unit, wherein the information differs from a recognition of the multiple objects.

14. The method according to claim 1 wherein the applying, by spanning elements of an iteration that differs from a first iteration, the dimension expansion process comprises assigning output values that are indicative of an identity of the relevant spanning elements of the iteration.

15. The method according to claim 1 wherein the applying, by spanning elements of an iteration that differs from a first iteration, the dimension expansion process comprises assigning output values that are indicative a history of dimension expansion processes until the iteration that differs from the first iteration.

16. The method according to claim 1 wherein each spanning element is associated with a subset of reference identifiers; and wherein the determining of the relevancy of each spanning elements of the iteration is based a relationship between the subset of the reference identifiers of the spanning element and an output of a last merge operation before the iteration.

17. The method according to claim 1 wherein an output of a dimension expansion process of an iteration is a multidimensional representation of the media unit that comprises media unit regions of interest that are associated with one or more expansion processes that generated the regions of interest.

18. The method according to claim 17 wherein a merge operation of the iteration comprises selecting a subgroup of media unit regions of interest based on a spatial relationship between the subgroup of multidimensional regions of interest.

19. The method according to claim 18 comprising applying a merge function on the subgroup of multidimensional regions of interest.

20. The method according to claim 18 comprising applying an intersection function on the subgroup of multidimensional regions of interest.

21. The method according to claim 17 wherein a merge operation of the iteration is based on an actual size of one or more multidimensional regions of interest.

22. The method according to claim 17 wherein a merge operation of the iteration is based on relationship between sizes of the multidimensional regions of interest.

23. The method according to claim 17 wherein a merge operation of the iteration is based on changes of the media unit regions of interest during at least the iteration and one or more previous iteration.

24. A non-transitory computer readable medium for low-power calculation of a signature, the non-transitory computer readable medium stores instructions for:
receiving or generating a media unit of multiple objects;
processing the media unit by performing multiple iterations, wherein for each iteration of at least some of the multiple iterations comprises applying, by spanning elements of the iteration, a dimension expansion process that is followed by a merge operation; and
wherein the applying of the dimension expansion process of the iteration comprises:
determining a relevancy of the spanning elements of the iteration;
completing the dimension expansion process by relevant spanning elements of the iteration and reducing a power consumption of irrelevant spanning elements until, at least, a completion of the applying of the dimension expansion process;

determining identifiers that are associated with significant portions of an output of the multiple iterations; and
providing a signature that comprises the identifiers and represents the multiple objects.

25. A signature generator that comprises:
an input that is configured to receive or generate a media unit of multiple objects;
a processor that is configured to process the media unit by performing multiple iterations, wherein for each iteration of at least some of the multiple iterations comprises applying, by spanning elements of the iteration, a dimension expansion process that is followed by a merge operation; and
wherein the applying of the dimension expansion process of the iteration comprises:
determining a relevancy of the spanning elements of the iteration;
completing the dimension expansion process by relevant spanning elements of the iteration and reducing a power consumption of irrelevant spanning elements until, at least, a completion

Assignees

• Cortica Ltd

Inventors

• Raichelgauz Igal
• Odinaev Karina

Classifications

• G06V10/454

  Integrating the filters into a hierarchical structure, e.g. convolutional neural networks [CNN] · CPC title

• G06V10/25

  Determination of region of interest [ROI] or a volume of interest [VOI] · CPC title

• G06F18/2155

  characterised by the incorporation of unlabelled data, e.g. multiple instance learning [MIL], semi-supervised techniques using expectation-maximisation [EM] or naïve labelling · CPC title

• G06F18/217

  Validation; Performance evaluation; Active pattern learning techniques · CPC title

• G06F18/211

  Selection of the most significant subset of features · CPC title