Low-power calculation of a signature of a media unit

1. A method for low-power calculation of a signature of a media unit, the method comprises:

calculating multiple attributes of segments of the media unit;

wherein the calculating comprises:

determining, by each spanning element of multiple spanning elements, a relevancy of the spanning element to the media unit to provide irrelevant spanning elements and relevant spanning elements;

reducing a power consumption of each irrelevant spanning element; and

completing the calculating of the multiple attributes of segments of the media unit by the relevant spanning elements;

determining identifiers that are associated with significant attributes out of the multiple attributes of segments of the media unit; and

providing a signature that comprises the identifiers and represents the media unit;

wherein each spanning element is associated with a subset of one or more reference identifiers; and wherein the determining of a relevancy of the spanning element to the media unit is based on a relationship between the subset and the identifiers related to the media unit.

2. The method according to claim 1 wherein each spanning element is associated with a subset of one or more reference identifiers; and wherein a spanning element is relevant to the media unit when the identifiers related to the media unit comprises each reference identifier of the subset.

3. The method according to claim 1 wherein different spanning elements are associated with different subsets of one or more reference identifiers, the different subsets of the of one or more reference identifiers represent different combinations of objects.

4. The method according to claim 1 wherein different spanning elements are associated with different combinations of object identifiers.

5. The method according to claim 4 comprising selecting candidates to combinations of object identifiers based on an occurrence of the object identifiers in signatures of test images.

6. The method according to claim 4 comprising selecting candidates to combinations of object identifiers to provide decorrelated combinations.

7. The method according to claim 4 comprising selecting candidates to combinations of object identifiers to cover multiple possible media units.

8. A method for low-power calculation of a signature of a media unit, the method comprises: calculating multiple attributes of segments of the media unit; wherein the calculating comprises: determining, by each spanning element of multiple spanning elements, a relevancy of the spanning element to the media unit to provide irrelevant spanning elements and relevant spanning elements; reducing a power consumption of each irrelevant spanning element; and completing the calculating of the multiple attributes of segments of the media unit by the relevant spanning elements; determining identifiers that are associated with significant attributes out of the multiple attributes of segments of the media unit; and providing a signature that comprises the identifiers and represents the media unit; wherein the calculating of the multiple attributes of segments of the media unit is executed in multiple iterations; and wherein each iteration is executed by spanning elements associated with the iteration; wherein the determining, by each spanning element of multiple spanning elements, of the relevancy of the spanning element to the media unit is executed per iteration.

9. The method according to claim 8 wherein the multiple iterations are preceded by a calculation of initial media unit attributes by one or more layers of a neural network.

10. A non-transitory computer readable medium for low-power calculation of a signature of a media unit by a group of spanning elements, the non-transitory computer readable medium stores instructions for:

calculating, multiple attributes of segments of the media unit; wherein the calculating comprises determining, by each spanning element of multiple spanning elements, a relevancy of the spanning element to the media unit to provide irrelevant spanning elements and relevant spanning elements; reducing a power consumption of each irrelevant spanning element; and completing the calculating of the multiple attributes of segments of the media unit by the relevant spanning elements;

determining identifiers that are associated with significant attributes out of the multiple attributes of segments of the media unit; and

providing a signature that comprises the identifiers and represents the media unit;

wherein each spanning element is associated with a subset of one or more reference identifiers; and wherein the determining of a relevancy of the spanning element to the media unit is based on a relationship between the subset and the identifiers related to the media unit.

11. The non-transitory computer readable medium according to claim 10 wherein each spanning element is associated with a subset of one or more reference identifiers; and wherein a spanning element is relevant to the media unit when the identifiers related to the media unit comprises each reference identifier of the subset.

12. The non-transitory computer readable medium according to claim 10 wherein different spanning elements are associated with different subsets of one or more reference identifiers, the different subsets represent different combinations of objects.

13. The non-transitory computer readable medium according to claim 10 wherein different spanning elements are associated with different combinations of object identifiers.

14. The non-transitory computer readable medium according to claim 13 that stores instructions for selecting candidates to combinations of object identifiers based on an occurrence of the object identifiers in signatures of test images.

15. The non-transitory computer readable medium according to claim 13 that stores instructions for selecting candidates to combinations of object identifiers to provide decorrelated combinations.

16. The non-transitory computer readable medium according to claim 13 that stores instructions for selecting candidates to combinations of object identifiers to cover multiple possible media units.

17. A non-transitory computer readable medium for low-power calculation of a signature of a media unit by a group of spanning elements, the non-transitory computer readable medium stores instructions for: calculating, multiple attributes of segments of the media unit; wherein the calculating comprises determining, by each spanning element of multiple spanning elements, a relevancy of the spanning element to the media unit to provide irrelevant spanning elements and relevant spanning elements; reducing a power consumption of each irrelevant spanning element; and completing the calculating of the multiple attributes of segments of the media unit by the relevant spanning elements; determining identifiers that are associated with significant attributes out of the multiple attributes of segments of the media unit; and providing a signature that comprises the identifiers and represents the media unit;

wherein the calculating of the multiple attributes of segments of the media unit is executed in multiple iterations; and wherein each iteration is executed by spanning elements associated with the iteration; wherein the determining, by each spanning element of multiple spanning elements, of the relevancy of the spanning element to the media unit is executed per iteration.

18. The non-transitory computer readable medium according to claim 17 wherein the multiple iterations are preceded by a calculation of initial media unit attributes by one or more layers of a neural network.

19. A signature generator that comprises:

a processor that is configured to calculate multiple attributes of segments of the media unit; wherein the calculating comprises determining, by each spanning element of the multiple spanning elements, a relevancy of the calculation element to the media unit to provide irrelevant spanning elements and relevant spanning elements; reducing a power consumption of each irrelevant spanning element; and completing the calculating of the multiple attributes of segments of the media unit by the relevant spanning elements;

an identifier unit that is configured to determine identifiers that are associated with significant attributes out of the multiple attributes of segments of the media unit; and

an output that is configured to provide a signature that comprises the identifiers and represents the media unit;

wherein each spanning element is associated with a subset of one or more reference identifiers; and wherein the determining of a relevancy of the calculation element to the media unit is based on a relationship between the subset and the identifiers related to the media unit.

20. The signature generator according to claim 19 wherein each spanning element is associated with a subset of one or more reference identifiers; and wherein a calculation element is relevant to the media unit when the identifiers related to the media unit comprises each reference identifier of the subset.

21. A signature generator that comprises: a processor that is configured to calculate multiple attributes of segments of the media unit; wherein the calculating comprises determining, by each spanning element of the multiple spanning elements, a relevancy of the calculation element to the media unit to provide irrelevant spanning elements and relevant spanning elements; reducing a power consumption of each irrelevant spanning element; and completing the calculating of the multiple attributes of segments of the media unit by the relevant spanning elements; an identifier unit that is configured to determine identifiers that are associated with significant attributes out of the multiple attributes of segments of the media unit; and an output that is configured to provide a signature that comprises the identifiers and represents the media unit;

wherein the calculating of the multiple attributes of segments of the media unit is executed in multiple iterations; and wherein each iteration is executed by calculation elements associated with the iteration; wherein the determining, by each spanning element of multiple spanning elements, of the relevancy of the calculation element to the media unit is executed per iteration.

22. The signature generator according to claim 21 wherein the multiple iterations are preceded by a calculation of initial media unit attributes by one or more layers of a neural network.