Recognizer of content of digital signals

• US 7,707,425 B2
• Filed: 11/04/2004
• Issued: 04/27/2010
• Est. Priority Date: 04/24/2001
• Status: Expired due to Fees

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Mihcak, M. Kivanc, Venkatesan, Ramarathnam
• Primary Examiner(s): Dinh, Minh
• Application Number:
  US10/980,917
• Publication Number:
  US 20050084103A1
• Time in Patent Office: 2,000 Days
• Field of Search: None
• US Class Current: 713/180
• CPC Class Codes: G06F 16/433 using audio data
• CPC Class Codes: G06F 16/634 Query by example, e.g. query by humming
• CPC Class Codes: G06F 16/683 using metadata automatically derived from the content
• CPC Class Codes: G10H 1/0058 Transmission between separate instruments or between individual components of a musical system G10H1/0083 takes precedence

Abstract

Described herein is a technology for facilitating the recognition of the content of digital signals. This abstract itself is not intended to limit the scope of this patent. The scope of the present invention is pointed out in the appending claims.

Claims

1. A computer-implemented method facilitating identification of a digital signal, the method comprising:

obtaining a digital signal;

deriving an identification value representative of the digital signal such that perceptually distinct digital signals result in identification values that are approximately independent of one another and perceptually same digital signals result in identical identification values, wherein the deriving comprises;

transforming the digital signal into a digital signal transform;

randomly dividing the digital signal transform into multiple chunks, each chunk containing signal data, wherein the dividing is carried out recursively to form hierarchical levels of overlapping chunks; and

averaging, for each of the chunks, the signal data to produce corresponding chunk averages;

generating, based in part on the chunk averages, an exponential distribution having multiple distinct quantization levels;

randomly rounding each of the chunk averages to one of the quantization levels to produce rounded values; and

hashing a composite of the rounded values; and

indexing the digital signal using the identification value.

2. A computer-implemented method as recited in claim 1, wherein the digital signals are selected from a group consisting of digital audio signals, digital image signals, digital video signals, and digital multimedia signals.

3. A computer-implemented method as recited in claim 1, wherein the digital signals are digital audio signals.

4. A computer-implemented method as recited in claim 1, wherein the identification value is a hash value.

5. A computer-implemented method as recited in claim 1 further comprising storing the identification value in association with the digital signal.

6. A computer-implemented method as recited in claim 1 further comprising watermarking the digital signal using, in part, the identification value to produce a watermarked digital signal.

7. A computer-implemented method as recited in claim 1 further comprising comparing the identification value with another identification value derived from another digital signal.

8. A computer-implemented method as recited in claim 1 further comprising comparing identification values of two digital signals to determine if such values substantially match.

9. A computer-implemented method as recited in claim 8 further comprising indicating whether such values substantially match.

10. A computer-readable storage medium having computer-executable instructions that, when executed by a computer, performs the method as recited in claim 1.

11. A computer-readable storage medium having stored thereon a data structure, comprising:

a first data field containing a digital signal;

a second data field derived from the first data field by deriving an identification value representative of the digital signal such that perceptually distinct digital signals result in identification values that are approximately independent of one another and perceptually same digital signals result in identical identification values, wherein the deriving comprises;

transforming the digital signal into a digital signal transform;

randomly dividing the digital signal transform into multiple chunks, each chunk containing signal data, wherein the dividing is carried out recursively to form hierarchical levels of overlapping chunks;

averaging, for each of the chunks, the signal data to produce corresponding chunk averages;

generating, based in part on the chunk averages, an exponential distribution having multiple distinct quantization levels;

randomly rounding each of the chunk averages to one of the quantization levels to produce rounded values; and

hashing a composite of the rounded values; and

a third data field functioning to delimit the end of the data structure;

wherein the second data field is the index of the first data field.

12. A computer-readable storage medium as recited in claim 11, wherein the digital signals are selected from a group consisting of digital audio signals, digital image signals, digital video signals, and digital multimedia signals.

13. A computer-implemented method facilitating identification of a digital signal, the method comprising:

obtaining two or more digital signal;

for each of the obtained digital signals, deriving an identification value representative of each of the obtained digital signals such that perceptually distinct digital signals result in identification values that are approximately independent of one another and perceptually same digital signals result in identical identification values, wherein the deriving comprises;

transforming each of the obtained digital signals into a digital signal transform;

randomly dividing each of the digital signal transforms into multiple chunks, each chunk containing signal data, wherein the dividing is carried out recursively to form hierarchical levels of overlapping chunks;

averaging, for each of the chunks of each of the digital signal transforms, the signal data to produce corresponding chunk averages;

generating, based in part on the chunk averages, an exponential distribution having multiple distinct quantization levels;

randomly rounding each of the chunk averages to one of the quantization levels to produce rounded values; and

hashing a composite of the rounded values;

comparing identification values of each of the obtained digital signals to determine if such values substantially match;

indicating whether such values substantially match; and

indexing each obtained digital signals based upon its associated identification value.

14. A computer-implemented method as recited in claim 13, wherein the digital signals are selected from a group consisting of digital audio signals, digital image signals, digital video signals, and digital multimedia signals.

15. A computer-implemented method as recited in claim 13, wherein each of the identification values is a hash value.

16. A computer-implemented method as recited in claim 13 further comprising storing the identification values in association with its associated digital signal.

17. A computer-implemented method facilitating identification of a digital signal, the method comprising:

obtaining a digital signal;

deriving an identification value representative of the digital signal based upon intrinsic characteristics of the digital signal, wherein identification values derived from perceptually distinct digital signals are approximately independent of one another and identification values derived from perceptually same digital signals are identical, wherein the deriving comprises;

transforming the digital signal into a digital signal transform;

randomly dividing the digital signal transform into multiple chunks, each chunk containing signal data, wherein the dividing is carried out recursively to form hierarchical levels of overlapping chunks; and

averaging, for each of the chunks, the signal data to produce corresponding chunk averages;

generating, based in part on the chunk averages, an exponential distribution having multiple distinct quantization levels;

randomly rounding each of the chunk averages to one of the quantization levels to produce rounded values; and

hashing a composite of the rounded values; and

indexing the digital signal using the identification value.

18. A computer-implemented method as recited in claim 17, wherein the digital signals are selected from a group consisting of digital audio signals, digital image signals, digital video signals, and digital multimedia signals.

19. A computer-implemented method as recited in claim 17, wherein the identification value is a hash value.

20. A computer-implemented method as recited in claim 17 further comprising storing the identification value in association with the digital signal.

21. A computer-implemented method as recited in claim 17 further comprising watermarking the digital signal using, in part, the identification value to produce a watermarked digital signal.

22. A computer-implemented method as recited in claim 17 further comprising comparing the identification value with another identification value derived from another digital signal.

23. A computer-implemented method as recited in claim 17 further comprising comparing identification values of two digital signals to determine if such values substantially match.

24. A computer-implemented method facilitating identification of a digital signal, the method comprising:

obtaining a digital signal;

deriving an identification value representative of the digital signal based upon intrinsic characteristics of the digital signal, wherein identification values derived from perceptually distinct digital signals do not substantially match one another and identification values derived from perceptually same digital signals substantially match one another, and the deriving comprises;

transforming the digital signal into a digital signal transform;

randomly dividing the digital signal transform into multiple chunks, each chunk containing signal data, wherein the dividing is carried out recursively to form hierarchical levels of overlapping chunks; and

averaging, for each of the chunks, the signal data to produce corresponding chunk averages;

generating, based in part on the chunk averages, an exponential distribution having multiple distinct quantization levels;

randomly rounding each of the chunk averages to one of the quantization levels to produce rounded values; and

hashing a composite of the rounded values; and

indexing the digital signal using the identification value.

25. A computer-implemented method as recited in claim 24, wherein the digital signals are selected from a group consisting of digital audio signals, digital image signals, digital video signals, and digital multimedia signals.

26. A computer-implemented method as recited in claim 24, wherein the identification value is a hash value.

27. A computer-implemented method as recited in claim 24 further comprising storing the identification value in association with the digital signal.

28. A computer-implemented method as recited in claim 24 further comprising watermarking the digital signal using, in part, the identification value to produce a watermarked digital signal.

29. A computer-implemented method as recited in claim 24 further comprising comparing the identification value with another identification value derived from another digital signal.

30. A computer-implemented method as recited in claim 24 further comprising comparing identification values of two digital signals to determine if such values substantially match.

31. One or more computer-readable storage media having stored thereon computer-executable instructions that, when executed by a computer, perform acts comprising:

obtaining a digital signal;

deriving an identification value representative of the digital signal based upon intrinsic characteristics of the digital signal, wherein identification values derived from perceptually distinct digital signals are approximately independent of one another and identification values derived from perceptually same digital signals are identical, and the deriving comprises;

transforming the digital signal into a digital signal transform;

randomly dividing the digital signal transform into multiple chunks, each chunk containing signal data, wherein the dividing is carried out recursively to form hierarchical levels of overlapping chunks;

averaging, for each of the chunks, the signal data to produce corresponding chunk averages;

generating, based in part on the chunk averages, an exponential distribution having multiple distinct quantization levels;

randomly rounding each of the chunk averages to one of the quantization levels to produce rounded values; and

hashing a composite of the rounded values; and

indexing the digital signal using the identification value.

32. One or more computer-readable storage media as recited in claim 31, wherein the digital signals are selected from a group consisting of digital audio signals, digital image signals, digital video signals, and digital multimedia signals.

33. One or more computer-readable storage media as recited in claim 31, wherein the identification value is a hash value.

34. One or more computer-readable storage media as recited in claim 31 further comprising storing the identification value in association with the digital signal.

35. One or more computer-readable storage media as recited in claim 31 further comprising watermarking the digital signal using, in part, the identification value to produce a watermarked digital signal.

36. One or more computer-readable storage media as recited in claim 31 further comprising comparing the identification value with another identification value derived from another digital signal.

37. One or more computer-readable storage media as recited in claim 31 further comprising comparing identification values of two digital signals to determine if such values substantially match.

38. A system comprising:

a processor;

a memory;

an acquisition means for obtaining a digital signal;

an identification-generation means for deriving an identification value representative of the digital signal based upon intrinsic characteristics of the digital signal, wherein identification values derived from perceptually distinct digital signals do not substantially match one another and identification values derived from perceptually same digital signals substantially match one another, wherein the deriving comprises;

transforming the digital signal into a digital signal transform;

randomly dividing the digital signal transform into multiple chunks, each chunk containing signal data, wherein the dividing is carried out recursively to form hierarchical levels of overlapping chunks; and

averaging, for each of the chunks, the signal data to produce corresponding chunk averages;

generating, based in part on the chunk averages, an exponential distribution having multiple distinct quantization levels;

randomly rounding each of the chunk averages to one of the quantization levels to produce rounded values; and

hashing a composite of the rounded values; and

an indexing means for indexing the digital signal using the identification value.

39. A system as recited in claim 38, wherein the digital signals are selected from a group consisting of digital audio signals, digital image signals, digital video signals, and digital multimedia signals.

40. A system as recited in claim 38, wherein the identification value is a hash value.

41. A system as recited in claim 38 further comprising a storage means for storing the identification value in association with the digital signal.

42. A system as recited in claim 38 further comprising a watermarking means for watermarking the digital signal using, in part, the identification value to produce a watermarked digital signal.

43. A system as recited in claim 38 further comprising a comparison means for comparing the identification value with another identification value derived from another digital signal.

44. A system as recited in claim 38 further comprising a comparison means for comparing identification values of two digital signals to determine if such values substantially match.