IP Library Granted Patent US 9,208,154
Granted Patent B1
US 9,208,154 · App. 14/458,387 · Granted Dec 8, 2015

IDF weighting of LSH bands for live reference ingestion

View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 9,208,154
App. No.
14/458,387
Granted
Dec 8, 2015
Kind
B1
Abstract

Down scoring overcrowded bands via IDF weighting scores provides a soft way to reduce the effect of common bands from Locality Sensitive Hashing (LSH) processes. An index component indexes live video references of a live streaming infrastructure pathway process in a reference index. A scoring component scores a set of bands with a set of inverse document frequency (IDF) weighting scores in the reference index. A high score is generated for bands that are featured in a small number of references and a low score is generated for bands featured in a high number of references.

Claims (54)

1. A system, comprising:

a memory that stores computer executable components;

a processor that executes the following computer executable components stored in the memory:

a scoring component that scores a set of bands via a set of Inverse Document Frequency (IDF) weighting scores in a media reference index,

wherein one or more live media references of a live streaming infrastructure process are indexed in the media reference index, and

wherein the scoring component generates a first IDF weighting score to a first set of bands that are featured in less than a first defined number of media references, and a second IDF weighting score to a second set of bands that are featured in more than a second defined number of media references.

2. The system of claim 1 , wherein the first IDF weighting score is greater than the second IDF weighting score.

3. The system of claim 1 , further comprising:

a reference component that:

identifies a probe video;

performs one or more lookups of Locality Sensitive Hashing (LSH) bands of the set of bands against the reference index from the probe video; and

determines a number of the media references having matching bands in which the first set of bands and the second set of bands are respectively featured.

4. The system of claim 3 , wherein the media references vary in length.

5. The system of claim 3 , further comprising:

a summing component that combines the set of IDF weighting scores from the matching bands within a media reference of the media references.

6. The system of claim 5 , further comprising:

a ranking component that ranks the media references according to the combined IDF weighting scores.

7. The system of claim 6 , further comprising:

a time alignment component that determines a sequence of matching offsets from the media references by applying a time alignment constraint on the matching bands.

8. The system of claim 1 , further comprising:

an offset component that determines a time offset for a band of a video reference; and

an index component that stores the band and corresponding time offset of the video reference in the media reference index.

9. The system of claim 1 , wherein the scoring component reduces a weight of references within the second set of bands that exceed a defined threshold by down scoring the second set of bands and, and wherein each band of the set of bands is associated with at least one of the set of IDF weighting scores.

10. The system of claim 1 , wherein the scoring component further computes the set of the IDF weighting scores based on a total number of offsets.

11. A computer-readable storage medium storing executable instructions that, when executed on a processor, cause the processor to perform operations comprising:

indexing media references in a reference index; and

scoring a set of bands via a set of inverse document frequency (IDF) weighting scores in the media reference index, wherein the scoring comprises:

generating a first IDF weighting score of the IDF weighting scores to a first set of bands of the set of bands that have band references in less than a first defined number of references; and

generating a second IDF weighting score of the IDF weighting scores to a second set of bands that have band references in more than a second defined number of media references.

12. The computer-readable storage medium of claim 11 , wherein the operations further comprise:

summing the IDF weighting scores from matching bands within the media references; and

ranking the media references according to the summed IDF weighting scores.

13. The computer-readable storage medium of claim 12 , wherein the operations further comprise:

applying one or more time alignments to the matching bands; and

determining a sequence of the media references that include rankings satisfying a defined threshold based on the one or more time alignments.

14. The computer-readable storage medium of claim 11 , wherein the first IDF weighting score is greater than the second IDF weighting score.

15. The computer-readable storage medium of claim 11 , wherein the operations further comprise:

identifying a probe video;

performing one or more lookups of locality sensitive hashing (LSH) bands of the set of bands against the index bin from the probe video; and

determining a number of the media references in which the first set of bands and the second set of bands are respectively featured.

16. The computer-readable storage medium of claim 11 , wherein the operations further comprise:

determining a time offset for the media references; and

computing the set of IDF weighting scores based on a total number of time offsets.

17. The computer-readable storage medium of claim 11 , wherein the scoring further comprises:

reducing a weight of entries within the subset of bands exceeding a defined threshold based on down scoring a subset of the set of bands via the set of IDF weighting scores.

18. A method comprising:

using a processor to execute computer executable instructions stored in a memory to perform operations comprising:

indexing live video references in a reference index within an live streaming infrastructure pathway; and

scoring a set of bands received in the live streaming infrastructure pathway via a set of inverse document frequency (IDF) weighting scores.

19. The method of claim 18 , further comprising:

storing the set of bands with the set of IDF weighting scores in a reference index, wherein the scoring comprises reducing a weight of entries within the subset of bands exceeding a defined threshold based on down scoring a subset of the set of bands via the IDF weighting scores.

20. The method of claim 18 , wherein the storing comprises:

generating a first IDF weighting score of the set of IDF weighting scores to a first set of bands of the set of bands that have band references in less than a first defined number of references; and

generating a second IDF weighting score of the set of IF weighting scores to a second set of bands that have band references in more than a second defined number of references.

Assignments (2)
CHANGE OF NAME Recorded Oct 2, 2017
From: GOOGLE INC.
To: GOOGLE LLC
Reel/Frame 044566/0657 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Aug 13, 2014
From: SHARIFI, MATTHEW; POSTELNICU, GHEORGHE
To: GOOGLE INC.
Reel/Frame 033523/0814 →