Content recognizer via probabilistic mirror distribution

• US 7,006,703 B2
• Filed: 06/28/2002
• Issued: 02/28/2006
• Est. Priority Date: 06/28/2002
• Status: Expired due to Fees

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Venkatesan, Ramarathnam, Johnson, Mark C.
• Primary Examiner(s): Patel, Kanjibhai
• Assistant Examiner(s): Bayat, Ali
• Application Number:
  US10/186,459
• Publication Number:
  US 20040005097A1
• Time in Patent Office: 1,341 Days
• Field of Search: 382/252, 382/171, 382/172, 382/173, 382/176, 382/180, 382/190, 382/220, 382/229, 382/240, 382/250, 382/251, 382/253, 382/300, 358/465, 358/466, 375/240, 375/240.25, 375/240.26, 375/240.03, 375/243
• US Class Current: 382/252
• CPC Class Codes: G06V 10/507 Summing image-intensity values; Histogram projection analysis

Abstract

An implementation of a technology, described herein, for facilitating the recognition of content of digital goods. At least one implementation, described herein, derives a probabilistic mirror distribution of a digital good (e.g., digital image or audio signal). It uses the resulting data to derive weighting factors (i.e., coefficients) for the digital good. Based, at least in part on such weighting factors, it determines statistics of the good and quantizes it. The scope of the present invention is pointed out in the appending claims.

Claims

1. A computer-readable medium having computer-executable instructions that, when executed by a computer, performs a method facilitating the recognition of content of digital goods, the method comprising:

obtaining a digital good;

segmenting the good into a plurality of regions;

for each region of the plurality;

defining one or more sub-regions inside of a region ;

computing suitable statistics for sub-region ;

quantizing the statistics for region ;

generating weighting factor m_ij for one or more points x_ij of region ;

producing a quantization q_R of region based upon one or more weighting factors m_ij of one or more points x_ij;

producing a recognition indication based upon a combination of the quantizations q_R of the regions of the plurality.

2. A medium as recited in claim 1, wherein a sub-region is centered on x_ij.

3. A medium as recited in claim 1, wherein x_ij is the center of region .

4. A medium as recited in claim 1, wherein a suitable statistic is a finite order moment of a region or sub-region.

5. A medium as recited in claim 1, wherein the segmenting generates multiple, pseudorandomly sized and distributed regions of the plurality.

6. A medium as recited in claim 1, wherein the generating weighting factor m_ij is generated in accordance with this formula:

m_ij⇓

N({tilde over (μ

)}_ij, σ

_ij)

where N is a normal distribution and σ

_ij is variance.

7. A medium as recited in claim 1, wherein the producing a quantization q_R is produced in accordance with this formula:

$q_R=\frac{\underset{\mathrm{ij}}{\sum <br><br>}<br><br>m_{\mathrm{ij}}<br><br>x_{\mathrm{ij}}}{\underset{\mathrm{ij}}{\sum <br><br>}<br><br>{\mathrm{\lambda <br><br>}}_{\mathrm{ij}}<br><br><br><br>x_{\mathrm{ij}}<br><br>}.$

8. A medium as recited in claim 7, wherein λ

_ij=1.

9. A medium as recited in claim 7, wherein λ

_ij pseudorandomly generated independent of the good.

10. A medium as recited in claim 1, wherein the producing a quantization q_R comprises hashing an unknown region M using x_ij as weights, where region M is based upon the weighting factor m_ij.

11. A device comprising a computer-readable medium as recited in claim 1.

12. A computer-readable medium having computer-executable instructions that, when executed by a computer, performs a method facilitating the recognition of content of digital goods, the method comprising:

obtaining a digital good;

segmenting the good into a plurality of regions;

for each region of the plurality;

generating weighting factor m_ij for one or more points x_ij of a region;

producing a quantization q_R of a region based upon one or more weighting factors m_ij of one or more points x_ij;

producing a recognition indication based on the quantizations q_R of the regions of the plurality.

13. A medium as recited in claim 12, wherein the generating comprises:

defining one or more sub-regions inside of a region ;

computing suitable statistics for sub-region ;

quantizing the statistics for region ;

generating weighting factor m_ij for one or more points x_ij of region .

14. A medium as recited in claim 13, wherein a sub-region is centered on x_ij.

15. A medium as recited in claim 13, wherein a suitable statistic is a finite order moment of a region or sub-region.

16. A medium as recited in claim 13, wherein the generating weighting factor m_ij generated in accordance with this formula:

m_ijN({tilde over (μ

)}_ij, σ

_ij)

where N is a normal distribution and σ

_ij is variance.

17. A medium as recited in claim 12, wherein x_ij is the center of region .

18. A medium as recited in claim 12, wherein the segmenting generates multiple, pseudorandomly sized and distributed regions of the plurality.

19. A medium as recited in claim 12, wherein the producing a quantization q_R is produced in accordance with this formula:

$q_R=\frac{\underset{\mathrm{ij}}{\sum <br><br>}<br><br>m_{\mathrm{ij}}<br><br>x_{\mathrm{ij}}}{\underset{\mathrm{ij}}{\sum <br><br>}<br><br>{\mathrm{\lambda <br><br>}}_{\mathrm{ij}}<br><br><br><br>x_{\mathrm{ij}}<br><br>}.$

20. A medium as recited in claim 12, wherein the producing a quantization q_R comprises hashing an unknown region M using x_ij as weights, where region M is based upon the weighting factor m_ij.

21. A device comprising a computer-readable medium as recited in claim 12.

22. A method facilitating the recognition of content of digital goods, the method comprising:

obtaining a digital good;

segmenting the good into a plurality of regions;

for each region of the plurality;

generating weighting factor m_ij for one or more points x_ij of a region;

producing a quantization q_R of a region based upon one or more weighting factors m_ij of one or more points x_ij;

producing a recognition indication based on the quantizations q_R of the regions of the plurality.

23. A method as recited in claim 22, wherein the generating comprises:

defining one or more sub-regions inside of a region ;

computing suitable statistics for sub-region ;

quantizing the statistics for region ;

generating weighting factor m_ij for one or more points x_ij of region .

24. A method as recited in claim 23, wherein a sub-region is centered on x_ij.

25. A method as recited in claim 23, wherein a suitable statistic is a finite order moment of a region or sub-region.

26. A method as recited in claim 23, wherein the generating weighting factor m_ij is generated in accordance with this formula:

m_ijN({tilde over (μ

)}_ij, σ

_ij)

where N is a normal distribution and σ

_ij is variance.

27. A method as recited in claim 22, wherein x_ij is the center of region .

28. A method as recited in claim 22, wherein the segmenting generates multiple, pseudorandomly sized and distributed regions of the plurality.

29. A method as recited in claim 22, wherein the producing a quantization q_R is produced in accordance with this formula:

$q_R=\frac{\underset{\mathrm{ij}}{\sum <br><br>}<br><br>m_{\mathrm{ij}}<br><br>x_{\mathrm{ij}}}{\underset{\mathrm{ij}}{\sum <br><br>}<br><br>{\mathrm{\lambda <br><br>}}_{\mathrm{ij}}<br><br><br><br>x_{\mathrm{ij}}<br><br>}.$

30. A method as recited in claim 22, wherein the producing a quantization q_R comprises hashing an unknown region M using x_ij as weights, where region M is based upon the weighting factor m_ij.

31. A computer-readable medium having data structure comprising:

one or more fields identifying a digital good, wherein such fields are selected from a group consisting of name, publishers, authors, licensing information, and serial number;

another field comprising a recognition indication associated with the digital good and the recognition indication produced by a method as recited in claim 22 based upon the digital good.

32. A system comprising:

a segmenter configured to generate multiple, pseudorandomly sized and distributed plurality of regions;

a weighting-factor generator configured to generate a weighting factor m_ij for one or more points x_ij of one or more regions of the plurality;

a region quantizer configured to produce a quantization q_R of a region based upon one or more weighting factors m_ij of one or more points x_ij of one or more regions of the plurality;

a combiner configured to produce a recognition indication based on the quantizations q_R of the regions of the plurality.

33. A system as recited in claim 32 further comprising a region transformer configured to:

define one or more sub-regions inside of a region ;

compute suitable statistics for sub-region ;

quantize the statistics for region .

34. A system as recited in claim 32, wherein the region quantizer is further configured to produce quantization q_R in accordance with this formula:

$q_R=\frac{\underset{\mathrm{ij}}{\sum <br><br>}<br><br>m_{\mathrm{ij}}<br><br>x_{\mathrm{ij}}}{\underset{\mathrm{ij}}{\sum <br><br>}<br><br>{\mathrm{\lambda <br><br>}}_{\mathrm{ij}}<br><br><br><br>x_{\mathrm{ij}}<br><br>}.$