HIERARCHICAL SAMPLING FOR OBJECT IDENTIFICATION

1 . A method of identifying targets, comprising:

receiving a first plurality of snapshots;

generating a first plurality of descriptors each associated with the first plurality of snapshots;

grouping the first plurality of snapshots into at least one cluster based on the plurality of descriptors;

selecting a representative snapshot for each of the at least one cluster;

generating at least one second descriptor for each representative snapshot, wherein the at least one second descriptor is more complex than the first plurality of descriptors; and

identifying a target based on comparing the at least second descriptor and a third descriptor.

2 . The method of claim 1 , wherein the third descriptor is associated with a second plurality of snapshots or an input query.

3 . The method of claim 2 , wherein the input query includes one or more arrays of numbers representing an intended target.

4 . The method of claim 1 , further comprising, prior to receiving the first plurality of snapshots:

receiving a second plurality of snapshots;

generating a third plurality of descriptors each associated with the second plurality of snapshots, wherein the third plurality of descriptors are less complex than the first plurality of descriptors;

grouping the second plurality of snapshots into a plurality of clusters based on the third plurality of descriptors; and

selecting a second plurality of representative snapshots as the first plurality of snapshots.

5 . The method of claim 1 , further comprising:

classifying the representative snapshot for each of the at least one cluster;

aggregating classification scores of the representative snapshot for each of the at least one cluster;

determining a class based on the aggregated classification scores; and

wherein the at least one descriptor is a class-specific descriptor.

6 . The method of claim 1 , further comprising, prior to selecting the representative snapshot, estimating a mean average precision (MAP) for each snapshot in the at least one cluster.

7 . The method of claim 6 , wherein selecting the representative snapshot for each of the at least one cluster comprises selecting a snapshot in the at least one cluster having a highest estimated MAP.

8 . The method of claim 6 , wherein estimating a MAP comprises using a neural network to estimate the MAP.

9 . A non-transitory computer readable medium comprising instructions stored therein that, when executed by a processor of a system, cause the processor to:

receive a first plurality of snapshots;

generate a first plurality of descriptors each associated with the first plurality of snapshots;

group the first plurality of snapshots into at least one cluster based on the plurality of descriptors;

select a representative snapshot for each of the at least one cluster;

generate at least one second descriptor for each representative snapshot, wherein the at least one second descriptor is more complex than the first plurality of descriptors; and

identify a target based on comparing the at least second descriptor and a third descriptor.

10 . The non-transitory computer readable medium of claim 9 , wherein the third descriptor is associated with a second plurality of snapshots or an input query.

11 . The non-transitory computer readable medium of claim 10 , wherein the input query includes one or more arrays of numbers representing an intended target.

12 . The non-transitory computer readable medium of claim 9 , further comprising instructions that, prior to receiving the first plurality of snapshots, cause the processor to:

receive a second plurality of snapshots;

generate a third plurality of descriptors each associated with the second plurality of snapshots, wherein the third plurality of descriptors are less complex than the first plurality of descriptors;

group the second plurality of snapshots into a plurality of clusters based on the third plurality of descriptors; and

select a second plurality of representative snapshots as the first plurality of snapshots.

13 . The non-transitory computer readable medium of claim 9 , further comprising instructions that cause the processor to:

classify the representative snapshot for each of the at least one cluster;

aggregate classification scores of the representative snapshot for each of the at least one cluster;

determine a class based on the aggregated classification scores; and

wherein the at least one descriptor is a class-specific descriptor.

14 . The non-transitory computer readable medium of claim 9 , further comprising instructions that, prior to selecting the representative snapshot, cause to processor to estimate a mean average precision (MAP) for each snapshot in the at least one cluster.

15 . The non-transitory computer readable medium of claim 14 , wherein the instructions for selecting the representative snapshot for each of the at least one cluster comprises instructions for selecting a snapshot in the at least one cluster having a highest estimated MAP.

16 . The non-transitory computer readable medium of claim 14 , wherein the instructions for estimating a MAP comprises instructions for using a neural network to estimate the MAP.

17 . A system, comprising:

memory that stores instructions; and

a processor configured to execute the instructions to:

receive a first plurality of snapshots;

generate a first plurality of descriptors each associated with the first plurality of snapshots;

group the first plurality of snapshots into at least one cluster based on the plurality of descriptors;

select a representative snapshot for each of the at least one cluster;

generate at least one second descriptor for each representative snapshot, wherein the at least one second descriptor is more complex than the first plurality of descriptors; and

identify a target based on comparing the at least second descriptor and a third descriptor.

18 . The system of claim 17 , wherein the third descriptor is associated with a second plurality of snapshots or an input query.

19 . The system of claim 18 , wherein the input query includes one or more arrays of numbers representing an intended target.

20 . The system of claim 17 , wherein the processor is further configured to, prior to selecting the representative snapshot, estimate a mean average precision (MAP) for each snapshot in the at least one cluster using a neural network.