System, method, and product for information embedding using an ensemble of non-intersecting embedding generators

• US 6,314,192 B1
• Filed: 05/21/1998
• Issued: 11/06/2001
• Est. Priority Date: 05/21/1998
• Status: Expired due to Term

• Current Assignee: Massachusetts Institute of Technology
• Original Assignee: Massachusetts Institute of Technology
• Inventors: Chen, Brian, Wornell, Gregory W.
• Primary Examiner(s): Johns, Andrew W.
• Application Number:
  US09/082,632
• Time in Patent Office: 1,265 Days
• Field of Search: 382/100, 382/232, 380/54, 380/210, 380/287, 380/252, 713/176, 370/527, 370/529
• US Class Current: 382/100
• CPC Class Codes: G06T 1/0028 Adaptive watermarking, e.g. Human Visual System [HVS]-based watermarking
• CPC Class Codes: G06T 1/0071 using multiple or alternating watermarks
• CPC Class Codes: G06T 2201/0202 whereby the quality of watermarked images is measured; Measuring quality or performance of watermarking methods; Balancing between quality and robustness
• CPC Class Codes: H04K 1/00 Secret communication
• CPC Class Codes: H04N 1/32208 involving changing the magnitude of selected pixels, e.g. overlay of information or super-imposition H04N1/32219, H04N1/32224 take precedence
• CPC Class Codes: H04N 1/32229 with selective or adaptive application of the additional information, e.g. in selected regions of the image H04N1/32213 takes precedence
• CPC Class Codes: H04N 1/32251 in multilevel data, e.g. greyscale or continuous tone data
• CPC Class Codes: H04N 1/32288 Multiple embedding, e.g. cocktail embedding, or redundant embedding, e.g. repeating the additional information at a plurality of locations in the image
• CPC Class Codes: H04N 1/32309 in colour image data

Abstract

A system, method, and product are provided to (1) embed a watermark signal into a host signal, thereby generating a composite signal, (2) optionally enable the composite signal to be transmitted over a communication channel, and (3) optionally extract the watermark signal from the transmitted composite signal. In one embodiment, the invention is a method for watermarking a host signal with a watermark signal. The watermark signal is made up of watermark-signal components, each having one of two or more watermark-signal values. The host signal is made up of host-signal components, each having one of two or more host-signal values. The method includes: (1) generating two or more embedding generators, each corresponding to a single watermark-signal value of a co-processed group of one or more watermark-signal components; (2) having each embedding generator generate two or more embedding values, the total of which is referred to as an original embedding-value set such that at least one embedding value generated by one embedding generator is different than any embedding value generated by another embedding generator; and (3) setting a host-signal value of one or more selected host-signal components to an embedding value of a particular embedding generator, thereby forming a composite-signal value, such that the particular embedding generator corresponds to the watermark-signal value of the co-processed group of watermark-signal components, and such that the embedding value of the particular embedding generator is selected based on its proximity to the host-signal value.

Claims

1. A method for watermarking a host signal with a watermark signal, the watermark signal comprising watermark-signal components, each having one of a plurality of watermark-signal values, and the host signal comprising host-signal components, each having one of a plurality of host-signal values, the method comprising:

(1) generating, by each of two or more of a plurality of embedding generators, each corresponding to a single watermark-signal value of one watermark-signal component or a co-processed group of two or more watermark-signal components, a plurality of embedding values, a total of each plurality of embedding values comprising a first embedding-value set, wherein at least one embedding value generated by at least one embedding generator is not the same as any embedding value generated by at least one other embedding generator; and

(2) setting at least one host-signal value of one or more selected host-signal components to a first embedding value of a first embedding generator, thereby forming a composite-signal value, wherein the first embedding generator corresponds to a first watermark-signal value of the one watermark-signal component or the co-processed group of watermark-signal components, and when each of the embedding generators is a quantizer, at least one quantization interval of at least one quantizer is not the same as any quantization interval of at least one other quantizer;

wherein, when each of the embedding generators is a dithered quantizer, each having quantization values that are uniformly spaced, and when the composite signal value is transmitted over a channel, then at least one quantization value of any of the dithered quantizers, plus at least one channel noise value capable of being induced by the channel, is not the same value as any of the quantization values of any of the dithered quantizers.

2. The method of claim 1, wherein:

at least one of the plurality of embedding generators is a quantizer.

3. The method of claim 2, wherein:

the at least one quantizer has a dimension that is greater than or equal to two.

4. The method of claim 3, wherein:

the at least one quantizer is a dithered quantizer.

5. The method of claim 2, wherein:

the at least one quantizer has a dimension that is equal to one.

6. The method of claim 5, wherein:

the at least one quantizer is a dithered quantizer.

7. The method of claim 1, wherein:

the first embedding value is an embedding value that is the closest of all embedding values of the first embedding generator in distance to the at least one host-signal value.

8. The method of claim 7, wherein:

the distance is determined by a Euclidean measure.

9. The method of claim 7, wherein:

the distance is determined by a non-uniformly weighted Euclidean measure.

10. The method of claim 7, wherein:

the distance is determined by a non-Euclidean measure.

11. The method of claim 1, wherein step 1 comprises:

generating the plurality of embedding values based on a maximum allowable watermark-induced distortion level and a maximum allowable channel-induced distortion level.

12. The method of claim 11, wherein the maximum allowable watermark-induced distortion level and the maximum allowable channel-induced distortion level have a relationship wherein when one increases the other does not decrease, and when one decreases, the other does not increase.

13. The method of claim 1, wherein step 1 comprises:

generating the plurality of embedding values based on a maximum allowable watermark-induced distortion level, a maximum allowable channel-induced distortion level, and a maximum allowable number of watermark-signal components per host-signal component.

14. The method of claim 13, wherein the maximum allowable watermark-induced distortion level and the maximum allowable channel-induced distortion level have a relationship wherein when one increases the other does not decrease, and when one decreases, the other does not increase.

15. The method of claim 1, wherein step 1 is based on a first predetermined relationship between each of the plurality of embedding values generated by at least one of the plurality of embedding generators.

16. The method of claim 15, wherein step 1 further is based on a second predetermined relationship between a second embedding value generated by one embedding generator and a third embedding value generated by a another embedding generator of the plurality of embedding generators.

17. The method of claim 16, wherein:

the second predetermined relationship is a not a dithered relationship.

18. The method of claim 16, wherein the second predetermined relationship is a dithered relationship.

19. The method of claim 18, wherein the dithered relationship is between quantized embedding values.

20. The method of claim 18, wherein the dithered relationship is between analog embedding values.

21. The method of claim 1, wherein step 1, is based on a first predetermined list including at least one of the plurality of embedding values generated by at least one of the plurality of embedding generators.

22. The method of claim 21, wherein step 1 further is based on a predetermined relationship between a second embedding value generated by the one embedding generator and a third embedding value generated by another embedding generator of the plurality of embedding generators.

23. The method of claim 22, wherein the predetermined relationship is a dithered relationship.

24. The method of claim 1, further comprising:

(3) extracting a reconstructed watermark-signal value from a received composite signal with noise value that includes the composite-signal value and a channel noise value, wherein the reconstructed watermark-signal value represents the first watermark-signal value.

25. The method of claim 24, wherein step 3 comprises:

(a) acquiring the received composite signal with noise value;

(b) replicating one or more embedding values from the first embedding-value set to form a second embedding-value set, at least one embedding value of the second embedding-value set having the same correspondence to a single watermark-signal value as have embedding values of the first embedding-value set from which the at least one embedding value of the second embedding value set is replicated;

(c) selecting a second embedding value of the second embedding-value set, wherein the second embedding value is selected based on its proximity to the received composite signal with noise value; and

(d) setting the reconstructed watermark-signal value to a one of the plurality of watermark-signal values to which the second embedding value corresponds.

26. The method of claim 25, wherein:

the second embedding value is an embedding value that is the closest of all embedding values of the second embedding-value set in distance to the received composite signal with noise value.

27. The method of claim 26, wherein:

the distance is determined by a Euclidean measure.

28. The method of claim 24, wherein the channel noise value is a value not including zero.

29. The method of claim 1, wherein step 1 comprises:

generating the plurality of embedding values based on a maximum allowable watermark-induced distortion level.

30. The method of claim 1, wherein:

each embedding value of the first embedding-value set is a uniquely mapped embedding value.

31. The method of claim 1, wherein:

the watermark signal is determined, at least in part based on the host signal.

32. The method of claim 1, wherein:

the watermark signal is error-correction coded.

33. The method of claim 1, wherein:

the watermark signal is error-detection coded.

34. The method of claim 1, wherein:

the watermark signal is encrypted.

35. The method of claim 1, wherein:

the watermark signal is a transformed signal.

36. The method of claim 1, wherein:

for at least one embedding generator, the embedding values are evenly spaced.

37. The method of claim 1, wherein:

for at least one embedding generator, the embedding values are unevenly spaced.

38. The method of claim 1, wherein:

at least one plurality of embedding values is generated by an embedding generator selected from one or more of the group consisting of a list, description, table, formula, or function.

39. The method of claim 1, wherein:

for at least one host-signal component, the plurality of host-signal values are scalar values.

40. The method of claim 1, wherein:

for at least one host-signal component, the plurality of host-signal values are vector values.

41. The method of claim 1, wherein:

for at least one host-signal component, the plurality of host-signal values represent values of a type selected from one of the group consisting of amplitude, phase, frequency, linear predictive coding coefficient, or nonlinear representation.

42. The method of claim 1, wherein:

for at least one watermark-signal component, the plurality of watermark-signal values are scalar values .

43. The method of claim 1, wherein:

for at least one watermark-signal component, the plurality of watermark-signal values are vector values.

44. The method of claim 1, wherein:

the first embedding value is selected based on its proximity to the at least one host-signal value.

45. The method of claim 1, wherein;

at least one embedding value of the first embedding-value set is predetermined.

46. The method of claim 1, wherein:

at least one embedding value of the first embedding-value set is not a quantization value.

47. The method of claim 1, wherein:

the watermark signal comprises an authentication signal.

48. The method of claim 1, wherein:

the watermark signal comprises a digital signature.

49. A method for extracting a reconstructed watermark-signal value from a received composite signal with noise value that includes a composite-signal value and a channel noise value, wherein the reconstructed watermark-signal value represents a first watermark-signal value of a plurality of watermark-signal values of one watermark-signal component or a co-processed group of two or more watermark-signal components of a watermark signal, wherein

the composite-signal value is formed by setting at least one host-signal value of one or more selected host-signal components of a host signal to a first embedding value of a first embedding generator that corresponds to the first watermark-signal value, and when each of the embedding generators is a quantizer, at least one quantization interval of at least one quantizer is not the same as any quantization interval of at least one other quantizer;

and further wherein the first embedding value is one of a first embedding-value set of embedding values generated by a plurality of embedding generators, each of two or more of which generate a plurality of embedding values, wherein at least one embedding value generated by at least one embedding generator of the plurality of embedding generators is not the same as any embedding value generated by at least one other embedding generator of the plurality of embedding generators;

the method comprising;

(1) acquiring the received composite signal with noise value;

(2) replicating one or more embedding values from the first embedding-value set to form a second embedding-value set, at least one embedding value of the second embedding-value set having the same correspondence to a single watermark-signal value as have embedding values of the first embedding-value set from which the at least one embedding value of the second embedding value set is replicated;

(3) selecting a second embedding value of the second embedding-value set, wherein the second embedding value is selected based on its proximity to the received composite signal with noise value; and

(4) setting the reconstructed watermark-signal value to a one of the plurality of watermark-signal values to which the second embedding value corresponds;

wherein, when each of the embedding generators is a dithered quantizer, each having quantization values that are uniformly spaced, and when the composite signal value is transmitted over a channel, then at least one quantization value of any of the dithered quantizers, plus at least one channel noise value capable of being induced by the channel, is not the same value as any of the quantization values of any of the dithered quantizers.

50. The method of claim 49, wherein:

the second embedding value is an embedding value that is the closest of all embedding values of the second embedding-value set in distance to the received composite signal with noise value.

51. The method of claim 50, wherein:

the distance is determined by a Euclidean measure.

52. The method of claim 49, wherein the channel noise value is a value not including zero.

53. The method of claim 49, wherein step (2) comprises:

replicating the one or more embedding values from the first embedding-value set based on at least a portion of the received composite signal with noise.

54. The method of claim 49, wherein step (2) comprises:

replicating the one or more embedding values from the first embedding-value set based on an externally provided specification of a first plurality of embedding values.

55. The method of claim 49, wherein:

the first embedding value is selected based on its proximity to the at least one host-signal value.

56. A system that watermarks a host signal with a watermark signal, the watermark signal comprising watermark-signal components, each having one of a plurality of watermark-signal values, and the host signal comprising host-signal components, each having one of a plurality of host-signal values, the system comprising:

an ensemble generator that generates a plurality of embedding generators, each corresponding to a single watermark-signal value of one watermark-signal component or a co-processed group of two or more watermark-signal components;

an embedding value generator that generates, by each of two or more of the plurality of embedding generators, a plurality of embedding values, the total of each plurality of embedding values comprising a first embedding-value set, wherein at least one embedding value generated by at least one embedding generator is not the same as any embedding value generated by at least one other embedding generator; and

a point coder that sets at least one host-signal value of one or more selected host-signal components to a first embedding value of a first embedding generator, thereby forming a composite-signal value, wherein the first embedding generator corresponds to a first watermark-signal value of the one watermark-signal component or the group of co-processed watermark-signal components, and, when each of the embedding generators is quantizer, at least one quantization interval of at least one quantizer is not the same as any quantization interval of at least one other quantizer,

wherein, when each of the embedding generators is a dithered quantizer, each having quantization values that are uniformly spaced, and when the composite signal value is transmitted over a channel, then at least one quantization value of any of the dithered quantizers, plus at least one channel noise value capable of being induced by the channel, is not the same value as any of the quantization values of any of the dithered quantizers.

57. The system of claim 56, wherein:

at least one of the plurality of embedding generators is a quantizer.

58. The system of claim 56, wherein:

the first embedding value is an embedding value that is the closest of all embedding values of the first embedding generator in distance to the at least one host-signal value.

59. The system of claim 56, wherein the ensemble generator generates the plurality of embedding generators based on a maximum allowable watermark-induced distortion level and a maximum allowable channel-induced distortion level.

60. The system of claim 59, wherein the maximum allowable watermark-induced distortion level and the maximum allowable channel-induced distortion level have a relationship wherein when one increases the other does not decrease, and when one decreases, the other does not increase.

61. The system of claim 56, wherein the ensemble generator generates the plurality of embedding generators based on a maximum allowable watermark-induced distortion level, a maximum allowable channel-induced distortion level, and a maximum allowable number of watermark-signal components per host-signal component.

62. The system of claim 61, wherein the maximum allowable watermark-induced distortion level and the maximum allowable channel-induced distortion level have a relationship wherein when one increases the other does not decrease, and when one decreases, the other does not increase.

63. The system of claim 56, wherein the embedding value generator generates a first plurality of embedding values based on a first predetermined relationship between each of the two or more embedding values generated by at least one of the plurality of embedding generators.

64. The system of claim 56, wherein the embedding value generator generates a first plurality of embedding values based on a second predetermined relationship between a second embedding value generated by one embedding generator and a third embedding value generated by another embedding generator of the plurality of embedding generators.

65. The system of claim 64, wherein the second predetermined relationship is a dithered relationship.

66. The system of claim 65, wherein the dithered relationship is between quantized embedding values.

67. The system of claim 66, wherein the dithered relationship is between analog embedding values.

68. The system of claim 56, wherein the embedding value generator generates a first plurality of embedding values based on a first predetermined list including at least one of the plurality of embedding values generated by at least one plurality of embedding generators.

69. The system of claim 68, wherein the embedding value generator further generates a first plurality of embedding values based on a predetermined relationship between a second embedding value generated by one embedding generator and a third embedding value generated by another embedding generator of the plurality of embedding generators.

70. The system of claim 69, wherein the predetermined relationship is a dithered relationship.

71. The system of claim 56, further comprising:

an information extractor that extracts a reconstructed watermark-signal value from a received composite signal with noise value that includes the composite-signal value and a channel noise value, wherein the reconstructed watermark-signal value represents the first watermark-signal value.

72. The system of claim 71, wherein the information extractor comprises:

a synchronizer that acquires a received composite signal with noise that includes the received composite signal with noise value;

an ensemble replicator that replicates one or more embedding values from the first embedding-value set to form a second embedding-value set, at least one embedding value of the second embedding-value set having the same correspondence to a single watermark-signal value as have embedding values of the first embedding-value set from which the at least one embedding value of the second embedding value set is replicated;

a point decoder that selects a second embedding value of the second embedding-value set based on its proximity to the received composite signal with noise value, and that sets the watermark-signal value to a one of the plurality of watermark-signal values to which the second embedding value corresponds.

73. The system of claim 72, wherein the synchronizer is further configured to determine the location of the received composite signal with noise value within the received composite signal with noise.

74. The system of claim 72, wherein the replication is based on an externally provided specification of the first plurality of embedding values.

75. The system of claim 72, wherein:

the second embedding value is an embedding value that is the closest of all embedding values of the second embedding-value set in distance to the received composite signal with noise value.

76. The system of claim 75, wherein:

the distance is determined by a Euclidean measure.

77. The system of claim 72, wherein the channel noise value is a value not including zero.

78. The system of claim 72, further comprising:

a transmitter to receive a composite signal including the composite-signal value from the point coder and provide the composite signal to a communication channel; and

a receiver to acquire the received composite signal with noise from the communication channel and provide it to the synchronizer.

79. The system of claim 72, wherein:

the ensemble replicator replicates the first embedding-value set based on at least a portion of the received composite signal with noise.

80. The system of claim 72, wherein:

the ensemble replicator replicates the one or more of the embedding values from the first embedding-value set based on an externally provided specification of a first plurality of embedding values.

81. The system of claim 56, wherein:

the one or more selected host-signal components are processed.

82. The system of claim 56, wherein:

the co-processed group of one or more watermark-signal components is processed.

83. The system of claim 56, wherein:

the host signal is a digital signal.

84. The system of claim 56, wherein:

the host signal is an analog signal.

85. The system of claim 56, wherein:

the watermark signal is a digital signal.

86. The system of claim 56, wherein:

the watermark signal is an analog signal.

87. The system of claim 56, further comprising:

a watermark-signal value determiner that selects the co-processed group of one or more watermark-signal components based on a maximum allowable channel-induced distortion level.

88. The system of claim 56, further comprising:

a watermark-signal value determiner that selects the co-processed group of one or more watermark-signal components based on a total number of watermark-signal components in the watermark signal.

89. The system of claim 56, wherein the ensemble generator generates the plurality of embedding generators based on a maximum allowable watermark-induced distortion level.

90. The system of claim 56, wherein:

each embedding value of the first embedding value set is a uniquely mapped embedding value.

91. The system of claim 56, wherein:

the watermark signal is determined, at least in part, based on the host signal.

92. The system of claim 56, wherein:

the watermark signal is error-correction coded.

93. The system of claim 56, wherein:

the watermark signal is error-detection coded.

94. The system of claim 56, wherein:

the watermark signal is encrypted.

95. The system of claim 56, wherein:

the watermark signal is a transformed signal.

96. The system of claim 56, wherein:

for at least one embedding generator, the embedding values are evenly spaced.

97. The system of claim 56, wherein:

for at least one embedding generator, the embedding values are unevenly spaced.

98. The system of claim 56, wherein:

at least one plurality of embedding values is generated by an embedding generator selected from one or more of the group consisting of a list, description, table, formula, or function.

99. The system of claim 56, wherein:

for at least one host-signal component, the plurality of host-signal values are scalar values.

100. The system of claim 56, wherein:

for at least one host-signal component, the plurality of host-signal values are vector values.

101. The system of claim 56, wherein:

for at least one host-signal component, the plurality of host-signal values represent values of a type selected from one of the group consisting of amplitude, phase, frequency, linear predictive coding coefficient, or nonlinear representation.

102. The system of claim 56, wherein:

for at least one watermark-signal component, the plurality of watermark-signal values are scalar values.

103. The system of claim 56, wherein:

for at least one watermark-signal component, the plurality of watermark-signal values are vector values.

104. The system of claim 56, wherein

the first embedding value is selected based on its proximity to the last one host-signal value.

105. The system of claim 56, further comprising:

a host signal analyzer and block selector that selects one or more host-signal components for embedding.

106. The system of claim 105, wherein:

at least one host-signal component of a co-processed group of host-signal components selected by the host signal analyzer and block selector is not quantized.

107. The system of claim 105, wherein:

the host signal analyzer and block selector selects one or more host-signal components based on their conveying important information.

108. The system of claim 105, wherein:

the host signal analyzer and block selector selects one or more host-signal components based on their not conveying important information.

109. A computer system that watermarks a host signal with a watermark signal, the watermark signal comprising watermark-signal components, each having one of a plurality of watermark-signal values, and the host signal comprising host-signal components, each having one of a plurality of host-signal values, the computer system comprising:

at least one embedding computer having an information embedder that embeds a watermark signal into a host signal, thereby creating a composite signal, the information embedder comprising;

an ensemble generator that generates a plurality of embedding generators, each corresponding to a single watermark-signal value of one watermark-signal component or a co-processed group of two or more watermark-signal components;

an embedding value generator that generates, by each of two or more of the plurality of embedding generators, a plurality of embedding values, the total of each plurality of embedding values comprising a first embedding-value set, wherein at least one embedding value generated by at least one embedding generator is not the same as any embedding value generated by at least one other embedding generator; and

a point coder that sets at least one host-signal value of one or more selected host-signal components to a first embedding value of a first embedding generator, thereby forming a composite-signal value, wherein the first embedding generator corresponds to a first watermark-signal value of the one watermark-signal component or the group of co-processed watermark-signal components, and, when each of the embedding generators is a quantizer, at least one quantization interval of at least one quantizer is not the same as any quantization interval of at least one other quantizer; and

at least one extracting computer having an information extractor that extracts the first watermark-signal value from the first embedding value;

wherein, when each of the embedding generators is a dithered quantizer, each having quantization values that are uniformly spaced, and when the composite signal value is transmitted over a channel, then at least one quantization value of any of the dithered quantizers, plus at least one channel noise value capable of being induced by the channel, is not the same value as any of the quantization values of any of the dithered quantizers.

110. The computer system of claim 109, wherein the embedding value generator generates a first plurality of embedding values based on a first predetermined relationship between each of the two or more embedding values generated by at least one of the plurality of embedding generators.

111. The computer system of claim 109, wherein the embedding value generator generates a first plurality of embedding values based on a second predetermined relationship between a second embedding value generated by one embedding generator and a third embedding value generated by another embedding generator of the plurality of embedding generators.

112. The computer system of claim 109, wherein the second predetermined relationship is a dithered relationship.

113. The computer system of claim 109, wherein the embedding value generator generates a first plurality of embedding values based on a first predetermined list including at least one of the plurality of embedding values generated by at least one of the plurality of embedding generators.

114. The computer system of claim 109, wherein the information extractor comprises:

a synchronizer that acquires a received composite signal with noise that includes the received composite signal with noise value;

an ensemble replicator that replicates one or more embedding values from the first embedding-value set to form a second embedding-value set, at least one embedding value of the second embedding-value set having the same correspondence to a single watermark-signal value as have the one embedding values of the first embedding-value set from which the at least one embedding value of the second embedding value set is replicated;

a point decoder that selects a second embedding value of the second embedding-value set based on its proximity to the received composite signal with noise value, and that sets the reconstructed watermark-signal value to a one of the plurality of watermark-signal values to which the second embedding value corresponds.

115. The computer system of claim 114, wherein:

the second embedding value is an embedding value that is the closest of all embedding values of the second embedding-value set in distance to the received composite signal with noise value.

116. The computer system of claim 109, wherein:

the embedding computer and the extracting computer are the same computer.

117. The computer system of claim 109, further comprising:

a transmitter coupled to the embedding computer to receive a composite signal including the composite-signal value from the point coder and provide the composite signal to a communication channel; and

a receiver coupled to the extracting computer to acquire the received composite signal with noise from the communication channel and provide it to the synchronizer.

118. The computer system of claim 109, wherein:

the first embedding value is selected based on its proximity to the at least one host-signal value.

119. A computer system that extracts a reconstructed watermark-signal value from a received composite signal with noise value that includes a composite-signal value and a channel noise value, wherein the reconstructed watermark-signal value represents a first watermark-signal value of a plurality of watermark-signal values of one watermark-signal component or a co-processed group of two or more watermark-signal components of a watermark signal, wherein

the composite-signal value is formed by setting at least one host-signal value of one or more selected host-signal components of a host signal to a first embedding value of a first embedding generator that corresponds to the first watermark-signal value, and, when each of the embedding generators is a quantizer, at least one quantization interval of at least one quantizer is not the same as any quantization interval of at least one other quantizer;

and further wherein the first embedding value is one of a first embedding-value set of embedding values generated by a plurality of embedding generators, each of two or more of which generate a plurality of embedding values, wherein at least one embedding value generated by at least one embedding generator of the plurality of embedding generators is not the same as any embedding value generated by at least one other embedding generator of the plurality of embedding generators;

the computer system comprising;

(a) a synchronizer that acquires the received composite signal with noise value;

(b) an ensemble replicator that replicates one or more embedding values from the first embedding-value set to form a second embedding-value set, at least one embedding value of the second embedding-value set having the same correspondence to a single watermark-signal value as have embedding values of the first embedding-value set from which the at least one embedding value of the second embedding value set is replicated;

(c) a point decoder that selects a second embedding value of the second embedding-value set, wherein the second embedding value is selected based on its proximity to the received composite signal with noise value, and sets the reconstructed watermark-signal value to a one of the plurality of watermark-signal values to which the second embedding value corresponds;

wherein, when each of the embedding generators is a dithered quantizer, each having quantization values that are uniformly spaced, and when the composite signal value is transmitted over a channel, then at least one quantization value of any of the dithered quantizers, plus at least one channel noise value capable of being induced by the channel, is not the same value as any of the quantization values of any of the dithered quantizers.

120. The computer system of claim 119, wherein:

the first embedding value is selected based on its proximity to the at least one host-signal value.

121. Storage media that contains software that, when executed on an appropriate computing system, performs a method for watermarking a host signal with a watermark signal, the watermark signal comprising watermark-signal components, each having one of a plurality of watermark-signal values, and the host signal comprising host-signal components, each having one of a plurality of host-signal values, the method comprising:

(1) generating, by each of two or more of a plurality of embedding generators, each corresponding to a single watermark-signal value of one watermark-signal component or a co-processed group of two or more watermark-signal components, a plurality of embedding values, the total of each plurality of embedding values comprising a first embedding-value set, wherein at least one embedding value generated by at least one embedding generator is not the same as any embedding value generated by at least one other embedding generator; and

(2) setting at least one host-signal value of one or more selected host-signal components to a first embedding value of a first embedding generator, thereby forming a composite-signal value, wherein the first embedding generator corresponds to a first watermark-signal value of the one watermark-signal component or the group of co-processed watermark-signal components, and, when each of the embedding generators is a quantizer, at least one quantization interval of at least one quantizer is not the same as any quantization interval of at least one other quantizer;

wherein, when each of the embedding generators is a dithered quantizer, each having quantization values that are uniformly spaced, and when the composite signal value is transmitted over a channel, then at least one quantization value of any of the dithered quantizers, plus at least one channel noise value capable of being induced by the channel, is not the same value as any of the quantization values of any of the dithered quantizers.

122. The method of claim 121, further comprising:

(3) extracting a reconstructed watermark-signal value from a received composite signal with noise value that includes the composite-signal value and a channel noise value, wherein the reconstructed watermark-signal value represents the first watermark-signal value.

123. The method of claim 122, wherein step 3 comprises:

(a) acquiring the received composite signal with noise value;

(b) replicating one or more embedding values from the first embedding-value set to form a second embedding-value set, at least one embedding value of the second embedding-value set having the same correspondence to a single watermark-signal value as have embedding values of the first embedding-value set from which the at least one embedding value of the second embedding value set is replicated;

(c) selecting a second embedding value of the second embedding-value set, wherein the second embedding value is an embedding value that is selected based on its proximity to the received composite signal with noise value; and

(d) setting the reconstructed watermark-signal value to a one of the plurality of watermark-signal values to which the second embedding value corresponds.

124. The storage media of claim 121, wherein:

the first embedding value is selected based on its proximity to the at least one host-signal value.

125. A computer program product for use with an appropriate computing system, the computer program product comprising a computer usable medium having embodied therein computer readable program code method steps for watermarking a host signal with a watermark signal, the watermark signal comprising watermark-signal components, each having one of a plurality of watermark-signal values, and the host signal comprising host-signal components, each having one of a plurality of host-signal values, the computer readable program code method steps comprising:

(1) generating, by each of two or more of a plurality of embedding generators, each corresponding to a single watermark-signal value of one watermark-signal component or a co-processed group of two or more watermark-signal components, a plurality of embedding values, the total of each plurality of embedding values comprising a first embedding-value set, wherein at least one embedding value generated by at least one embedding generator is not the same as any embedding value generated by at least one other embedding generator; and

(2) setting at least one host-signal value of one or more selected host-signal components to a first embedding value of a first embedding generator, thereby forming a composite-signal value, wherein the first embedding generator corresponds to a first watermark-signal value of the one watermark-signal component or the group of co-processed watermark-signal components, and, when each of the embedding generators is a quantizer, at least one quantization interval of at least one quantizer is not the same as any quantization interval of at least one other quantizer;

wherein, when each of the embedding generators is a dithered quantizer, each having quantization values that are uniformly spaced, and when the composite signal value is transmitted over a channel, then at least one quantization value of any of the dithered quantizers, plus at least one channel noise value capable of being induced by the channel, is not the same value as any of the quantization values of any of the dithered quantizers.

126. The computer program product of claim 125, wherein the computer readable program code method steps further comprise:

(3) extracting a reconstructed watermark-signal value from a received composite signal with noise value that includes the composite-signal value and a channel noise values wherein the reconstructed watermark-signal value represents the first watermark-signal value.

127. The computer program product of claim 126, wherein step 3 comprises:

(a) acquiring the received composite signal with noise value;

(b) replicating one or more embedding values from the first embedding-value set to form a second embedding-value set, at least one embedding value of the second embedding-value set having the same correspondence to a single watermark-signal value as have embedding values of the first embedding-value set from which the at least one embedding value of the second embedding value set is replicated;

(c) selecting a second embedding value of the second embedding-value set, wherein the second embedding value is an embedding value that is selected based on its proximity to the received composite signal with noise value; and

(d) setting the reconstructed watermark-signal value to a one of the plurality of watermark-signal values to which the second embedding value corresponds.

128. The computer program product of claim 125, wherein:

the first embedding value is selected based on its proximity to the at least one host-signal value.

129. A method for watermarking a host signal with a watermark signal, the watermark signal comprising watermark-signal components, each having one of a plurality of watermark-signal values, and the host signal comprising host-signal components, each having one of a plurality of host-signal values, the method comprising:

(1) designating a plurality of embedding generators, each corresponding to a single watermark-signal value of one watermark-signal component or a co-processed group of two or more watermark-signal components, based at least in part on a maximum allowable channel-induced distortion level;

(2) generating, by each of two or more of the plurality of embedding generators, a plurality of embedding values, a total of each plurality of embedding values comprising a first embedding-value set, wherein at least one embedding value generated by at least one embedding generator is not the same as any embedding value generated by at least one other embedding generator; and

(3) setting at least one host-signal value of one or more selected host-signal components to a first embedding value of a first embedding generator, thereby forming a composite-signal value, wherein the first embedding generator corresponds to a first watermark-signal value of the one watermark-signal component or the co-processed group of watermark-signal components, and, when each of the embedding generators is a quantizer, at least one quantization interval of at least one quantizer is not the same as any quantization interval of at least one other quantizer.

130. The method of claim 129, wherein:

step (1) is further based at least in part on a maximum allowable watermark-induced distortion level and a maximum allowable number of watermark-signal components per host-signal component.

131. The method of claim 130, wherein:

the maximum allowable number of watermark-signal components per host-signal component is greater than or equal to two.

132. A system that watermarks a host signal with a watermark signal, the watermark signal comprising watermark-signal components, each having one of a plurality of watermark-signal values, and the host signal comprising host-signal components, each having one of a plurality of host-signal values, the system comprising:

an ensemble generator that generates a plurality of embedding generators, each corresponding to a single watermark-signal value of one watermark-signal component or a co-processed group of two or more watermark-signal components, wherein the ensemble generator generates the plurality of embedding generators based at least in part on a maximum allowable channel-induced distortion level;

an embedding value generator that generates, by each of two or more of the plurality of embedding generators, a plurality of embedding values, the total of each plurality of embedding values comprising a first embedding-value set, wherein at least one embedding value generated by at least one embedding generator is not the same as any embedding value generated by at least one other embedding generator; and

a point coder that sets at least one host-signal value of one or more selected host-signal components to a first embedding value of a first embedding generator, thereby forming a composite-signal value, wherein the first embedding generator corresponds to a first watermark-signal value of the one watermark-signal component or the group of co-processed watermark-signal components, and, when each of the embedding generators is quantizer, at least one quantization interval of at least one quantizer is not the same as any quantization interval of at least one other quantizer.

133. The system of claim 132, wherein:

the ensemble generator further generates the embedding generators based at least in part on a maximum allowable watermark-induced distortion level and a maximum allowable number of watermark-signal components per host-signal component.

the first embedding value is selected based on its proximity to the at least one host-signal value.

134. A method for watermarking a host signal with a watermark signal, the watermark signal comprising watermark-signal components, each having one of a plurality of watermark-signal values, and the host signal comprising host-signal components, each having one of a plurality of host-signal values, the method comprising:

(1) accessing a plurality of value-sets of two or more embedding values, the embedding values of each value-set corresponding to a single watermark-signal value of one watermark-signal component or a co-processed group of two or more watermark-signal components, wherein at least one embedding value of at least one value-set is not the same as any embedding value of at least one other value-set; and

(2) setting at least one host-signal value of one or more selected host-signal components to a first embedding value of a first value-set, thereby forming a composite-signal value, wherein the first value-set corresponds to a first watermark-signal value of the one watermark-signal component or the co-processed group of watermark-signal components, and, when the embedding values are quantized values, at least one quantization interval of at least one value-set is not the same as any quantization interval of at least one other value-set;

wherein, when each embedding value of each value-set is a dithered quantization value, when the embedding values of each value-set are uniformly spaced with respect to each other, and when the composite signal value is transmitted over a channel, then at least one quantization value of any of the dithered quantizers, plus at least one channel noise value capable of being induced by the channel, is not the same value as any of the quantization values of any of the dithered quantizers.

135. The method of claim 134, further comprising:

(3) extracting a reconstructed watermark-signal value from a received composite signal with noise value that includes the composite-signal value and a channel noise value, wherein the reconstructed watermark-signal value represents the first watermark-signal value.

136. The method of claim 135, wherein:

the channel noise value is a value not including zero.

137. A method for extracting a reconstructed watermark-signal value from a received composite signal with noise value that includes a composite-signal value and a channel noise value, wherein (a) the reconstructed watermark-signal value represents a first watermark-signal value of a plurality of watermark-signal values of one watermark-signal component or a co-processed group of two or more watermark-signal components of a watermark signal, (b) the composite-signal value is formed by accessing a plurality of value-sets of two or more embedding values, the embedding values of each value-set corresponding to a single watermark-signal value, wherein at least one embedding value of at least one value-set is not the same as any embedding value of at least one other value-set, and setting at least one host-signal value of one or more selected host-signal components to a first embedding value of a first value-set, thereby forming the composite-signal value, (c) the first value-set corresponds to a first watermark-signal value of the one watermark-signal component or the co-processed group of watermark-signal components, (d) when the embedding values are quantized values, at least one quantization interval of at least one value-set is not the same as any quantization interval of at least one other value-set, and (e) when each of the embedding generators is a dithered quantizer, each having quantization values that are uniformly spaced, and when the composite signal value is transmitted over a channel, then at least one quantization value of any of the dithered quantizers, plus at least one channel noise value capable of being induced by the channel, is not the same value as any of the quantization values of any of the dithered quantizers,

the method comprising:

(1) acquiring the received composite signal with noise value;

(2) replicating one or more embedding values from the first embedding-value set to form a second embedding-value set, at least one embedding value of the second embedding-value set having the same correspondence to a single watermark-signal value as have embedding values of the first embedding-value set from which the at least one embedding value of the second embedding value set is replicated;

(3) selecting a second embedding value of the second embedding-value set, wherein the second embedding value is selected based on its proximity to the received composite signal with noise value; and

(4) setting the reconstructed watermark-signal value to a one of the plurality of watermark-signal values to which the second embedding value corresponds.

138. A system for watermarking a host signal with a watermark signal, the watermark signal comprising watermark-signal components, each having one of a plurality of watermark-signal values, and the host signal comprising host-signal components, each having one of a plurality of host-signal values, wherein the system operates upon a plurality of value-sets of two or more embedding values, the embedding values of each value-set corresponding to a single watermark-signal value of one watermark-signal component or a co-processed group of two or more watermark-signal components, wherein at least one embedding value of at least one value-set is not the same as any embedding value of at least one other value-set the system comprising:

a point coder that sets at least one host-signal value of one or more selected host-signal components to a first embedding value of a first value-set, thereby forming a composite-signal value, wherein the first value-set corresponds to a first watermark-signal value of the one watermark-signal component or the co-processed group of watermark-signal components, and, when the embedding values are quantized values, at least one quantization interval of at least one value-set is not the same as any quantization interval of at least one other value-set;

wherein, when each of the embedding generators is a dithered quantizer, each having quantization values that are uniformly spaced, and when the composite signal value is transmitted over a channel, then at least one quantization value of any of the dithered quantizers, plus at least one channel noise value capable of being induced by the channel, is not the same value as any of the quantization values of any of the dithered quantizers.

139. The system of claim 138, further comprising:

an information extractor that extracts a reconstructed watermark-signal value from a received composite signal with noise value that includes the composite-signal value and a channel noise value, wherein the reconstructed watermark-signal value represents the first watermark-signal value.

140. The system of claim 139, wherein:

the channel noise value is a value not including zero.

141. A system for extracting a reconstructed watermark-signal value from a received composite signal with noise value that includes a composite-signal value and a channel noise value, wherein (a) the reconstructed watermark-signal value represents a first watermark-signal value of a plurality of watermark-signal values of one watermark-signal component or a co-processed group of two or more watermark-signal components of a watermark signal, (b) the composite-signal value is formed by accessing a plurality of value-sets of two or more embedding values, the embedding values of each value-set corresponding to a single watermark-signal value, wherein at least one embedding value of at least one value-set is not the same as any embedding value of at least one other value-set, and setting at least one host-signal value of one or more selected host-signal components to a first embedding value of a first value-set, thereby forming the composite-signal value, (c) the first value-set corresponds to a first watermark-signal value of the one watermark-signal component or the co-processed group of watermark-signal components, (d) when the embedding values are quantized values, at least one quantization interval of at least one value-set is not the same as any quantization interval of at least one other value-set, and (e) when each of the embedding generators is a dithered quantizer, each having quantization values that are uniformly spaced, and when the composite signal value is transmitted over a channel, then at least one quantization value of any of the dithered quantizers, plus at least one channel noise value capable of being induced by the channel, is not the same value as any of the quantization values of any of the dithered quantizers,

the system comprising an information extractor that extracts a reconstructed watermark-signal value from a received composite signal with noise value that includes the composite-signal value and a channel noise value, wherein the reconstructed watermark-signal value represents the first watermark-signal value.

142. The system of claim 141, wherein:

the information extractor comprises

(1) a synchronizer that acquires the received composite signal with noise value;

(2) an ensemble replicator that replicates one or more embedding values from the first embedding-value set to form a second embedding-value set, at least one embedding value of the second embedding-value set having the same correspondence to a single watermark-signal value as have embedding values of the first embedding-value set from which the at least one embedding value of the second embedding value set is replicated; and

(3) a point decoder that selects a second embedding value of the second embedding-value set, wherein the second embedding value is selected based on its proximity to the received composite signal with noise value, and that sets the reconstructed watermark-signal value to a one of the plurality of watermark-signal values to which the second embedding value corresponds.

143. A computer program product for use with an appropriate computing system, the computer program product comprising a computer usable medium having embodied therein computer readable program code method steps for extracting a reconstructed watermark-signal value from a received composite signal with noise value that includes a composite-signal value and a channel noise value, wherein (a) the reconstructed watermark-signal value represents a first watermark-signal value of a plurality of watermark-signal values of one watermark-signal component or a co-processed group of two or more watermark-signal components of a watermark signal, (b) the composite-signal value is formed by accessing a plurality of value-sets of two or more embedding values, the embedding values of each value-set corresponding to a single watermark-signal value, wherein at least one embedding value of at least one value-set is not the same as any embedding value of at least one other value-set, and setting at least one host-signal value of one or more selected host-signal components to a first embedding value of a first value-set, thereby forming the composite-signal value, (c) the first value-set corresponds to a first watermark-signal value of the one watermark-signal component or the co-processed group of watermark-signal components, (d) when the embedding values are quantized values, at least one quantization interval of at least one value-set is not the same as any quantization interval of at least one other value-set, and (e) when each of the embedding generators is a dithered quantizer, each having quantization values that are uniformly spaced, and when the composite signal value is transmitted over a channel, then at least one quantization value of any of the dithered quantizers, plus at least one channel noise value capable of being induced by the channel, is not the same value as any of the quantization values of any of the dithered quantizers,

the computer readable program code method steps comprising extracting a reconstructed watermark-signal value from a received composite signal with noise value that includes the composite-signal value and a channel noise value, wherein the reconstructed watermark-signal value represents the first watermark-signal value.

144. The computer program product of claim 143, wherein:

the extracting comprises

(a) acquiring the received composite signal with noise value;

(b) replicating one or more embedding values from the first embedding-value set to form a second embedding-value set, at least one embedding value of the second embedding-value set having the same correspondence to a single watermark-signal value as have embedding values of the first embedding-value set from which the at least one embedding value of the second embedding value set is replicated; and

(c) selecting a second embedding value of the second embedding-value set, wherein the second embedding value is selected based on its proximity to the received composite signal with noise value, and that sets the reconstructed watermark-signal value to a one of the plurality of watermark-signal values to which the second embedding value corresponds.