CA2487151C - Digital goods representation based upon matrix invariances - Google Patents

Digital goods representation based upon matrix invariances Download PDF

Info

Publication number
CA2487151C
CA2487151C CA2487151A CA2487151A CA2487151C CA 2487151 C CA2487151 C CA 2487151C CA 2487151 A CA2487151 A CA 2487151A CA 2487151 A CA2487151 A CA 2487151A CA 2487151 C CA2487151 C CA 2487151C
Authority
CA
Canada
Prior art keywords
pseudo
digital
regions
randomly
recited
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Expired - Fee Related
Application number
CA2487151A
Other languages
French (fr)
Other versions
CA2487151A1 (en
Inventor
Mehmet Kivanc Mihcak
Ramarathnam Venkatesan
Suleyman Serdar Kozat
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Microsoft Corp
Original Assignee
Microsoft Corp
Microsoft Technology Licensing LLC
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Microsoft Corp, Microsoft Technology Licensing LLC filed Critical Microsoft Corp
Publication of CA2487151A1 publication Critical patent/CA2487151A1/en
Application granted granted Critical
Publication of CA2487151C publication Critical patent/CA2487151C/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/50Information retrieval; Database structures therefor; File system structures therefor of still image data
    • G06F16/58Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
    • G06F16/583Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
    • G06F16/5854Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content using shape and object relationship
    • GPHYSICS
    • G10MUSICAL INSTRUMENTS; ACOUSTICS
    • G10LSPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
    • G10L15/00Speech recognition
    • G10L15/02Feature extraction for speech recognition; Selection of recognition unit
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N21/00Selective content distribution, e.g. interactive television or video on demand [VOD]
    • H04N21/80Generation or processing of content or additional data by content creator independently of the distribution process; Content per se
    • H04N21/83Generation or processing of protective or descriptive data associated with content; Content structuring
    • H04N21/835Generation of protective data, e.g. certificates
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F1/00Details not covered by groups G06F3/00 - G06F13/00 and G06F21/00
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/40Information retrieval; Database structures therefor; File system structures therefor of multimedia data, e.g. slideshows comprising image and additional audio data
    • G06F16/41Indexing; Data structures therefor; Storage structures
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/50Information retrieval; Database structures therefor; File system structures therefor of still image data
    • G06F16/58Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
    • G06F16/583Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
    • G06F16/5838Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content using colour
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06VIMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00Arrangements for image or video recognition or understanding
    • G06V10/40Extraction of image or video features
    • G06V10/42Global feature extraction by analysis of the whole pattern, e.g. using frequency domain transformations or autocorrelation
    • G06V10/431Frequency domain transformation; Autocorrelation

Abstract

Described herein is an implementation that produces a new representation of a digital good (such as an image) in a new defined representation domain. In particular, the representations in this new domain are based upon matrix invariances. In some implementations, the matrix invariances may, for example, heavily use singular value decomposition (SVD).

Description

DIGITAL GOODS REPRESENTATION BASED UPON MATRIX
IN S
TECHNICAL FIELD
[0001] This invention generally relates to a signal representation technology.

BACKGROUND
[0002] Digital goods are often distributed to consumers over private and public networks¨such as Intranets and the Internet. In addition, these goods are distributed to consumers via fixed computer readable media, such as a compact disc (CD-ROM), digital versatile disc (DVD), soft magnetic diskette, or hard magnetic disk (e.g., a preloaded hard drive).
[0003] Unfortunately, it is relatively easy for a person to pirate the pristine digital content of a digital good at the expense and harm of the content owners¨
which includes the content author, publisher, developer, distributor, etc. The content-based industries (e.g., entertainment, music, film, software, etc.) that produce and distribute content are plagued by lost revenues due to digital piracy.
[0004] "Digital goods" is a generic label, used herein, for electronically stored or transmitted content. Examples of digital goods include images, audio clips, video, multimedia, software, and data. Depending upon the context, digital goods may also be called a "digital signal," "content signal," "digital bitstream,"
"media signal," "digital object," "object," "signal," and the like.
[0005] In addition, digital goods are often stored in massive databases¨
either structured or unstructured. As these databases grow, the need for streamlined categorization and identification of goods increases.

Hashing
[0006] Hashing techniques are employed for many purposes. Among those purposes are protecting the rights of content owners and speeding database searching/access. Hashing techniques are used in many areas such as database management, querying, cryptography, and many other fields involving large amounts of raw data.
[0007] In general, a hashing technique maps a large block of raw data into a relatively small and structured set of identifiers. These identifiers are also referred to as "hash values" or simply "hash." By introducing a specific structure and order into raw data, the hashing function drastically reduces the size of the raw data into a smaller (and typically more manageable) representation.

Limitations of Conventional Hashing
[0008] Conventional hashing techniques are used for many kinds of data.
These techniques have good characteristics and are well understood.
Unfortunately, digital goods with visual and/or audio content present a unique set of challenges not experienced in other digital data. This is primarily due to the unique fact that the content of such goods is subject to perceptual evaluation by human observers. Typically, perceptual evaluation is visual and/or auditory.
[0009] For example, assume that the content of two digital goods is, in fact, different, but only perceptually, insubstantially so. A human observer may consider the content of two digital goods to be similar. However, even perceptually insubstantial differences in content properties (such as color, pitch, intensity, phase) between two digital goods result in the two goods appearing substantially different in the digital domain.
[0010] Thus, when using conventional hashing functions, a slightly shifted version of a digital good generates a very different hash value as compared to that of the original digital good, even though the digital good is essentially identical (i.e., perceptually the same) to the human observer.
[0011] The human observer is rather tolerant of certain changes in digital goods. For instance, human ears are less sensitive to changes in some ranges of frequency components of an audio signal than other ranges of frequency components.
[0012] This human tolerance can be exploited for illegal or unscrupulous purposes. For example, a pirate may use advanced audio processing techniques to remove copyright notices or embedded watermarks from audio signal without perceptually altering the audio quality.
[0013] Such malicious changes to the digital goods are referred to as "attacks", and result in changes at the data domain. Unfortunately, the human observer is unable to perceive these changes, allowing the pirate to successfully distribute unauthorized copies in an unlawful manner.
[0014] Although the human observer is tolerant of such minor (i.e., imperceptible) alterations, the digital observer¨in the form of a conventional hashing technique¨is not tolerant. Traditional hashing techniques are of little help identifying the common content of an original digital good and a pirated copy of such good because the original and the pirated copy hash to very different hash values. This is true even though both are perceptually identical (i.e., appear to be the same to the human observer).

Applications for Hashing Techniques
[0015] There are many and varied applications for hashing techniques.
Some include anti-piracy, content categorization, content recognition, watermarking, content-based key generation, and synchronization in audio or video streams.
[0016] Hashing techniques may be used to search on the Web for digital goods suspected of having been pirated. In addition, hashing techniques are used to generate keys based upon the content of a signal. These keys are used instead of or in addition to secret keys. Also, hashing functions may be used to synchronize input signals. Examples of such signals include video or multimedia signals. A hashing technique must be fast if synchronization is performed in real time.

SUMMARY
[0017] Described herein is an implementation that produces a new representation of a digital good (such as an image) in a new defined representation domain. In particular, the representations in this new domain are based upon matrix invariances. In some implementations, the matrix invariances may, for example, heavily use singular value decomposition (SVD).

[0017a] According to an aspect of the present invention, there is provided a method for producing a new representation of a digital good in a new defined representation domain, the new defined representation domain based upon matrix invariances, the method comprising: obtaining a digital goods; selecting, by a partitioner in a computing device, a plurality of pseudo-randomly sized and pseudo-randomly positioned regions from the digital goods; extracting robust features from the plurality of pseudo-randomly sized and pseudo-randomly positioned regions, wherein the features are based upon singular value decomposition (SVD), discrete cosine transform, or discrete wavelet transform and further wherein the features are within the new defined representation domain; producing a first output comprising the calculated statistics of one or more of the plurality of pseudo-randomly sized and pseudo-randomly positioned regions, wherein the statistics of the one or more of the plurality of pseudo-randomly sized and pseudo-randomly positioned regions are representatives of respective one or more of the plurality of pseudo-randomly sized and pseudo-randomly positioned regions and are calculated based upon matrix invariances; constructing a secondary representation of the digital goods by using a pseudo-random combination of the calculated statistics; forming multiple second regions from the secondary representation; generating a new set of feature vectors from each second region via a SVD transformation; and producing a second output.

[0017b] According to another aspect of the present invention, there is provided a system comprising: a processor; one or more memories having stored therein processor-executable modules, the modules comprising: an obtainer configured to obtain a digital goods; a partitioner configured to pseudo-randomly partition the digital goods into a plurality of regions based at least upon a secret key; a calculator configured to calculate singular vectors for one or more of the plurality of regions via singular value decompositions, wherein the singular vectors remain invariant with high probability for disturbances to the digital goods in a probability space defined by the secret key.
[0017c] According to still another aspect of the present invention, there is provided a non-transitory computer storage medium having processor-executable instructions that, when executed by a processor, performs a method comprising:

obtaining a digital goods; pseudo-randomly partitioning the digital goods into a plurality of regions; generating singular vectors for one or more of the plurality of regions via singular value decompositions, wherein the singular vectors remain invariant with high probability for disturbances to the digital goods.
[0017d] According to yet another aspect of the present invention, there is provided a computing device comprising: an output, which is configured to produce output that is audio, visual, or both; and a non-transitory medium as described above.
[0017e] According to a further aspect of the present invention, there is provided a non-transitory computer storage medium having processor-executable instructions that, when executed by a processor, performs a method facilitating protection of digital goods, the method comprising: obtaining a digital goods; partitioning the digital goods into a plurality of pseudo-randomly sized and pseudo-randomly positioned regions, after the partitioning the digital goods, calculating statistics of one or more of the plurality of pseudo-randomly sized and pseudo-randomly positioned regions, wherein the statistics of the one or more of the plurality of pseudo-randomly sized and pseudo-randomly positioned regions are representatives of respective one or more of the plurality of pseudo-randomly sized and pseudo-randomly positioned regions and are calculated based upon matrix invariances; producing output comprising the calculated statistics of the one or more of the plurality of pseudo-randomly sized and pseudo-randomly positioned regions.

5a = 51018-116 [00171 According to still a further aspect of the present invention, there is provided a processor-readable medium having processor-executable instructions that, when executed by a processor, performs a method for generating a representation of digital goods, the method comprising: obtaining a digital good;
segmenting the good into a plurality of regions; generating feature vectors from each of the regions of the plurality, wherein the feature vectors are calculated based upon matrix invariant singular value decomposition, SVD; and producing output using a combination of the calculated feature vectors, the output forming a hash vector for the digital good.

[0017g] According to another aspect of the present invention, there is provided a computing device comprising: an output, which is configured to produce output that is audio, visual, or both; and a non-transitory medium as described above.

[0017h] According to still another aspect of the present invention, there is provided a computer comprising one or more processor-readable media as described above.

[0017i] According to yet another aspect of the present invention, there is provided a method for generating a representation of digital goods, the method comprising: obtaining a digital good; segmenting the good into a plurality of regions;
extracting robust feature vectors from each of the regions of the plurality, wherein the feature vectors are calculated based upon matrix invariant singular value decomposition, SVD; and producing output using a combination of the calculated feature vectors, the output forming a hash vector for the digital good.

[0017j] According to a further aspect of the present invention, there is provided a system comprising: a processor; one or more memories having stored therein processor-executable modules, the modules comprising: an obtainer configured to obtain a digital good; a partitioner configured to segment the good into a plurality of regions; and a calculator configured to generate feature vectors from each of the regions of the plurality, wherein the feature vectors are calculated based upon matrix invariant singular value decomposition, SVD; and an output device configured to 5b = = 51018-116 produce output using a combination of the calculated feature vectors, the output forming a hash vector for the digital good.
BRIEF DESCRIPTION OF THE DRAWINGS
[0018] The same numbers are used throughout the drawings to reference like elements and features.
[0019] Fig. 1 is a flow diagram showing a methodological implementation described herein.
[0020] Fig. 2 is a block diagram of an implementation described herein.
[0021] Fig. 3 is an example of a computing operating environment capable of (wholly or partially) implementing at least one embodiment described herein.
DETAILED DESCRIPTION
[0022] In the following description, for purposes of explanation, specific numbers, materials and configurations are set forth in order to provide a thorough understanding of the present invention. However, it will be apparent to one skilled in the art that the present invention may be practiced without the specific exemplary details. In other instances, well-known features are omitted or 5c simplified to clarify the description of the exemplary implementations of the present invention, and thereby, to better explain the present invention.
Furthermore, for ease of understanding, certain method steps are delineated as separate steps; however, these separately delineated steps should not be construed as necessarily order dependent in their performance.
[0023] The following description sets forth one or more exemplary implementations of a Digital Goods Representation based upon Matrix Invariances that incorporate elements recited in the appended claims. These implementations are described with specificity in order to meet statutory written description, enabling, and best-mode requirements. However, the description itself is not intended to limit the scope of this patent.
[0024] These exemplary implementations, described herein, are examples.
These exemplary implementations do not limit the scope of the claimed present invention; rather, the present invention might also be embodied and implemented in other ways, in conjunction with other present or future technologies.
[0025] An example of an embodiment of a Digital Goods Representation based upon Matrix Invariances may be referred to as an "exemplary goods representer."
[0026] When randomization is mentioned herein, it should be understood that the randomization is carried out by means of a pseudo-random number generator (e.g., RC4) whose seed is the secret key (K), where this key is unknown to the adversary.

Introduction
[0027] The one or more exemplary implementations of the invention, described herein, may be implemented (wholly or partially) on computing systems and computer networks like that show in Fig. 3. Although implementations may have many applications, cryptosystems, authorization, and security are examples of particular applications.
[0028] The exemplary goods representer derives robust feature vectors of digital goods from pseudo-randomly selected semi-global regions of the goods via matrix invariances. Such regions may (but need not be) overlapping.
[0029] Unlike the conventional approaches, the exemplary goods representer's calculations are based on matrix invariances (such as that based upon Singular Value Decomposition (SVD)). SVD components capture essential characteristics of digital goods.

Semi-global Characteristics
[0030] Semi-global characteristics are representative of general characteristics of a group or collection of individual elements. As an example, they may be statistics or features of "regions" (i.e., "segments"). Semi-global characteristics are not representatives of the individual local characteristics of the individual elements; rather, they are representatives of the perceptual content of the group (e.g., segments) as a whole.
[0031] The semi-global characteristics may be determined by a mathematical or statistical representation of a group. For example, it may be an average of the color values of all pixels in a group. Consequently, such semi-global characteristics may also be called "statistical characteristics." Local characteristics do not represent robust statistical characteristics.

Notations
[0032] Herein, capital letters (e.g., A, B, C) represent matrices, lowercase letters with vector notation (e.g., ¨a, -b, ¨c) represent column vectors and lowercase letters represent scalars (e.g., a, b, c). The secret key is represented with K.
[0033] Herein, the following mathematical definitions are used:
= / Rnxn : Two-dimensional representation of digital goods of size n x n.
= 1,, Identity matrix of size n x n.
= Ai Elr" matrix which represents the ith pseudo-random region (e.g., a rectangle of size mxm) taken from the digital goods.
= AT Transpose of matrix A.
= 'Alp : The Frobenous norm of a matrix A defined as IA = (Ekm iErm where aki is the element of A at row k and column 1.
= AH : Hermitian transpose of matrix A. Note that AH= AT for real matrices.

= 1-V12 The L2 norm of a vector which is defined as I¨v12 = (k v)2 where vk is1 the k-th element of ¨v.
= D E le" size m DCT transformation matrix for 1-dimensional signals of length m. Note that 2-dimensional DCT transform of a matrix /(size m x m) is given by D/DT.
= W E Rm xm : size m DWT transformation matrix for 1-dimensional signals of length m. Note that 2-dimensional DWT transform of a matrix I (size m x m) is given by WIWT.
= Har: Hamming weight of a binary vector ¨a.
= SVD of a matrix A, el r" is defined as: A= ()EV H where ¨ U= [-441-142 = = = -447]{~ u, are orthonormal eigenvectors of the matrix AAH (and may not be unique in general). {¨ u1}71 are termed as the left singular vectors of A.
¨ V = . {¨ v1}11 are orthonormal eigenvectors of the matrix AHA (and may not be unique in general). {¨ v,}71 are termed as the right singular vectors of A.
¨>2: A diagonal real matrix of size m x m where the i-th diagonal entry, is termed as the i-th singular value. Without loss of generality, one may assume ai>= cs2...>= am.

Singular Value Decomposition (SVD)
[0034] The exemplary goods representer captures the essence of the geometric information while having dimensionality reduction. SVD has some provable optimality properties: "Best" lower-dimensional (say K-dimensional) approximation to a matrix (say rank N, N>= K) in the sense of Frobenius norm is produced by the first K singular vectors and the corresponding singular values.
[0035] The essence of the semi-global features and the geometric information of digital goods (such as images) are compactly captured by the significant components of the SVD of such goods. Such components are approximately invariant under intentional or unintentional disturbances as long as the digital goods of interest are not perceptively altered too severely.
[0036] With the exemplary goods representer, SVD is applied to pseudo-randomly-chosen semi-global regions of images mainly because of security reasons.
SVD components obtained from these regions accurately represent the overall features of the digital goods and bear favorable robustness properties while providing reasonable security as long as we use the sufficient number and size of regions.
[0037] The conventional choices were typically DCT (discrete cosine transform) and DWT (discrete wavelet transform). With DCT or DWT, the digital goods are projected onto a fixed set of fixed basis vectors. DCT/DWT have proven to be generally effective for conventional goods processing applications.
[0038] Instead of the DCT/DWT-type fixed basis transforms, the exemplary goods representer employs Singular Value Decomposition (SVD). With SVD, the exemplary goods representer selects the optimal basis vectors in L2 norm sense (see Equation (1) below). Furthermore, given a matrix, its SVD is unique. As an analogy, if a digital good is represented as a vector in some high-dimensional vector space, then the singular vectors give the optimal direction information to the good in the sense of Equation (1) while the singular values give the distance information along this direction. Consequently, the singular vectors that correspond to large singular vectors are naturally prone to any scaling attack and other small conventional signal-processing modifications.
[0039] By using SVD decomposition, the digital goods may be considered as a two dimensional surface in a three dimensional space. When DCT-like transformations are applied to a digital good (or surface), the information about any particularly distinctive (hence important) geometric feature of the digital good is dispersed to all coefficients.
[0040] As an example, an image may have a surface with strong peaks (e.g., very bright patches in a dark background) which will be dispersed to all transformations in case of DCT. By using SVD, the exemplary goods representer preserves both the magnitude of these important features (in singular values) and also their location and geometry in the singular vectors. Hence, the combination of the top left and right singular vectors (i.e. the ones that correspond to the largest singular values) captures the important geometric features in an image in L2 norm sense.

Properties of SVD
[0041] The following describes the mathematically properties of SVD.
Let A= n vH be the SVD of A. Then,
[0042] 1) The left singular vectors U = [ . .
¨um]: {¨u,}7_, are an orthonormal basis for the column space of A.
[0043] 2) The right singular vectors V = . .
{¨ v,}7_1 are an orthonormal basis for the row space of A.
[0044] 3) We have u,,¨ v,) = arg min IA¨a¨ x¨ y" 12, , where Hx12 = hyl2 = 1 and V k 1 < k<=m k-1 (ak lik Vk ) arg min I A ¨Zo-iu, 1=-1 ¨axy11 12, , (1) where cyj > = 02... > = cym are the singular values, {¨lid and {¨vd are the corresponding singular vectors.

Hashing
[0045] A hash function employed by the exemplary goods representer has two inputs, a digital good (such as an image) / and a secret key K. This hash function produces a short vector 4/ = HK (I) from a set {0, 1}" with 2"
cardinality.

It is desirable for the perceptual hash to be equal for all perceptual-similar digital goods with high probability. It is also desirable for two perceptually different digital goods to produce unrelated hash values with high probability. Such a hash function is a many-to-one mapping. On the other hand, for most applications it may be enough to have sufficiently similar (respectively different) hash values for perceptually similar (respectively different) inputs with high probability, i.e., the hash function may show a graceful change.
[0046] The requirements for such a hash function are given as:
[0047] 1) Randomization : For any given input, its hash value should be approximately uniformly distributed among all possible outputs. The probability measure is defined by the secret key.
[0048] 2) Pairwise Independence : The hash outputs for two perceptually different digital goods should be independent with high probability, where the probability space is defined by the secret key.
[0049] 3) Invariance : For all possible acceptable disturbances, the output of the hash function should remain approximately invariant with high probability, where the probability space is defined by the secret key.
[0050] Two digital goods are deemed to be perceptually similar when there are no reasonably noticeable distortions between them in terms of human perception.

Methodological Implementations of the Exemplary Goods Representer
[0051] Fig. 1 shows a methodological implementation of the exemplary goods representer. This methodological implementation may be performed in software, hardware, or a combination thereof
[0052] At 110, the exemplary goods representer obtains input digital goods.
For this explanation, the input digital goods will be an image of size n x n, which may be described as I E lrixn . Note that, the image may also be rectangular (i.e., the sizes may be different). This approach can be generalized to this condition with no difficulty.
[0053] At 120, the exemplary goods representer pseudo-randomly forms multiple regions from I. The number of regions may be called p and the shape of the regions may be, for example, rectangles. The shape of the regions may differ from implementation to implementation.
[0054] Although they do not necessarily need to, these regions may overlap each other. However, one may produce an implementation that requires such overlap. Conversely, one may produce an implementation that does not allow overlap.
[0055] This action, represented by: Ai E Rrnxrn, l<= i <= p. Ai is a matrix which represents the ith pseudo-random region (e.g., a rectangle of size mxm) taken from the digital goods. Note that, each of these regions can be a matrix of different sizes and this can be easily used in this approach with no difficulty.
[0056] At 130, it generates feature vectors (each of which may be labeled -gi) from each region Ai via a SVD-based transformation. This feature-vector generation may be generically described as --gi=
[0057] These feature vectors (-gi) may be used as hash values after suitable quantization or they can be used as intermediate features from which actual hash values may be produced. The SVD-based transformation (T/(Ai)) is a hash function that employs SVD. Examples of hash functions are described below in the section titled "SVD-based Hash Functions."
[0058] At this point, the exemplary goods representer has produced a representation (the collection of feature vectors produced by -gi = TAA,)) of the digital goods. Some implementations may end here with a combination of to form the hash vector.
[0059] In these implementations, T1() may be designed so that T1(A) yields the top q singular values from the rectangle Ai. Another possibility would be to design T1() such that Ti(A) yields the top q singular vectors (left, right or both).
These are the q singular vectors that correspond to the largest q values.
Naturally, in both cases, the parameter q should be chosen properly; for instance, a logical decision would require q <<m.
[0060] In some implementations, it would be possible to choose p= 1 and Ai such that it corresponds to the whole image. Note that this variant does not possess any randomness; hence, it is more suitable for non-adversarial applications of image hashing.
[0061] Alternatively, other implementations may perform additional processing to produce even smoother results. Blocks 140, 150, 160, and 170 show that.
[0062] At 140, the exemplary goods representer constructs a secondary representation J of the digital goods by using a pseudo-random combination of feature vectors {-gb At this point, these vectors produced as part of block 130 may be considered "intermediate" feature vectors.
[0063] As part of such construction of the secondary representation J, the exemplary goods representer collects the first left and right singular vectors that correspond to the largest singular value from each subsection.
[0064] Let F = where (respectively v) is the first left (respectively right) singular vector of the i-th subsection. Then, the exemplary goods representer pseudo-randomly forms a smooth representation J
from the set F: Given a pseudo-randomly selected initial singular vector, we proceed to form Jby selecting and replacing subsequent vectors from F such that the next chosen vector is closest to the previous vector in /.2norm sense.
[0065] Hence, after 2p steps all the elements of F are pseudo-randomly re-ordered and J(of size m x 2p) is formed. Note that, the 112 metric can be replaced by any other suitable metric (possibly randomized) in the formation of Jso that continuity and smoothness are achieved. The smooth nature ofJmay be desirable in some implementations.
[0066] Also note that, instead of this simple pseudo-random re-ordering of vectors, it is possible to apply other (possibly more complex) operations to generate J.
[0067] At 150, the exemplary goods representer pseudo-randomly forms multiple regions from J. The number of regions may be called r and the shape of the regions may be, for example, rectangles. The shape of the regions may differ from implementation to implementation. Like the above-described regions, these regions may be any shape and may overlap (but are not required to do so).
[0068] This action is represented by this: B, c Rdxd, .1<= i <= r. 13, is a matrix which represents the ith pseudo-random region (e.g., a rectangle of size dxd) taken from the secondary representation J of the digital goods. Note that, in this implementation, the rectangles may have different sizes. In other implementations, the rectangles may be the same size.
[0069] At 160, it generates a new set of feature vectors (each of which may be labeled ¨1) from each region Bi via a SVD-based transformation. This feature-vector generation may be generically described as -1,=T2(B,).
[0070] These feature vectors (-17) are hash values. The SVD-based transformation (T2(B1)) is a hash function that employs SVD. Examples of hash functions are described below in the section titled "SVD-based Hash Functions."
These two SVD-based transformations (T1 and T2) may be the same as or different from each other.
[0071] At 170, the exemplary goods representer combines the feature vectors of this new set --fpl to form the new hash vector, which produces an output that includes the combination of vectors.

SVD-based Hash Functions
[0072] This section discusses several hashing functions that may be employed by the SVD-based transformations (T1 and T2) introduced above in the description of Fig. 1.

SVD-SVD Hash Functions
[0073] Given an image, for example, the exemplary goods representer pseudo-randomly selects p subimages Ai E Rmxm, 1<= I <= p. Then the exemplary goods representer finds the SVD of each sub-image:
Ai = UiSiViT , where U, Vi are the m xm real left and right singular vector matrices respectively and Si is the real m x m diagonal matrix consisting of the singular values along the diagonal.
[0074] After forming the secondary representation at block 140, the exemplary goods representer reapplies the SVD to subsections of Bi's. As the hash vector, the exemplary goods representer keeps the corresponding set of the first rleft and right singular vectors from each Bi.after suitable quantization DCT-SVD
[0075] As a variant of the SVD-SVD approach, the exemplary goods representer uses 2D-DCT transform as the initial transformation (T1) in the block 130. After finding 2D-DCT of each sub-image Ai, DAiDT , only the top - band of frequencies from the coefficient matrix Di is preserved.
Here, D denotes the DCT transform matrix. The selection of fmin and fmax, determines the selected frequency band. The coefficients of low-to-mid band frequencies are more descriptive and distinctive for images. Selecting fmin >

avoids near DC frequencies, which are more sensitive to simple scaling or DC
level changes. Selecting a small value offma, avoids using coefficients of higher frequenciesõ which can be altered by small noise addition, smoothing, compression, etc. Hence, depending on the problem specifications, suitable values offmin and f,,a, can be chosen.
[0076] The coefficients in this frequency band are then stored as a vector E Rf*fn"x-fmµn*fm'n for each region Ai. The ordering of the elements of (a} is user-dependent and can possibly be used to introduce extra randomness. Then, a secondary representation is formed, following along the same lines, by choosing random vectors from the set F ={-d1,.. , ¨4)), and pseudo-randomly forming a smooth representation J Next, the exemplary goods representer applies SVD to 1.=
J= USVT , and stores the first left and right singular vectors ¨/41 and ¨v1 as the hash vectors.

DWT-SVD
100771 This is a variant of the DCT-SVD approach where the 2D-DCT is replaced with 2D-DWT. After getting random rectangles Ai's from the image, 1-level of DWT is applied to each A. The DC subbands are stored as vectors -di e R1'32 121 to form the secondary representation J in the next stage. Next, we apply SVD to J.-J= US VT
[0078] The first left and right singular vectors -441 and ¨v1 corresponding to the largest singular value are stored as the hash vectors after suitable quantization.

Binary SVD
100791 Instead of working on the original domain, the exemplary goods representer forms a binary representation from the original image, preserving significant regions of the digital goods. If the goods are an image, this approach might threshold the image pixels, where the threshold level is chosen, such that only t percent of image pixels are represented as ones (or zeros).
Alternatively, the threshold level can be chosen such that, in each subimage, only t percent of image pixels are ones (or zeros).
[0080] Given image /õ a binary image, after thresholding, may be represented as 'band first left and right binary singular vectors may be defined to correspond to the largest singular value as Hibl5Vb1)=. argmini/bE6' x YT IH
where --x and -y binary vectors and 9 is the binary xor operation. The other singular vectors may be found alternatively, such that the (ic + 1)-th singular vector pairs are derived from 'b Eik Ubl b7; ,k> 1 and 9 is for summation.
100811 Hence, after thresholding, the first binary singular vectors for each binary subimage is found and forms the set F =
, = = . ,---vbp). After forming the secondary binary representation .4 in the second stage, the exemplary goods representer proceeds by using the binary SVD on the r pseudo-randomly chosen regions. The final hash value is given by --12 =
(-fib ¨Ujn¨/Vjl, = = - 1-10-Direct SVD
[0082] T1 may be used as the identity transform and use the subsections directly. This idea is readily applicable to binary digital goods (such as a binary image 4)which can be formed after thresholding. From each subsection Ai of size m x m, form vectors -di e le' directly from the samples of the goods. The secondary representation Jis generated directly from F =

, -4). Next, the exemplary goods representer applies SVD to J
J = US VT
and stores the first left and right singular vectors --u1 and --v1 as the hash vectors.

Exemplary System for Generating Representation of Digital Goods [0083] Fig. 2 shows an exemplary system 200 for generating representation of digital goods, which is an example of an embodiment of the exemplary goods representer.
[0084] The system 200 generates a representation (e.g., a hash value) of a digital good. In this example, digital good is an image. The system 200 includes a goods obtainer 210, a partitioner 220, a region-statistics calculator 230, and an output device 240.
[0085] The goods obtainer 210 obtains a digital good 205 (such as an audio signal or a digital image). It may obtain the goods from nearly any source, such as a storage device or over a network communications link. In addition to obtaining, the goods obtainer 410 may also normalize the amplitude of the goods. In that case, it may also be called an amplitude normalizer.
[0086] The partitioner 220 separates the goods into multiple, pseudo-randomly sized, pseudo-randomly positioned regions (i.e., partitions). Such regions may overlap (but such overlap is not necessary).
[0087] For example, if the good is an image, it might be partitioned into two-dimensional polygons (e.g., regions) of pseudo-random size and location.
In another example, if the good is an audio signal, a two-dimensional representation (using frequency and time) of the audio clip might be separated into two-dimensional polygons (e.g., triangles) of pseudo-random size and location.

[0088] In this implementation, the regions may indeed overlap with each other.
[0089] For each region, the region-statistics calculator 230 calculates statistics of the multiple regions generated by the partitioner 220.
Statistics for each region are calculated. The statistics calculated by the calculator 230 may be the feature vectors described above in the description of blocks 130 and 160.
[0090] The output device 240 may present the results (for each region or combined) of the region-statistics calculator 230. Such results may be stored or used for further calculations.

Examples of Applications for Exemplary Goods Representer [0091] The exemplary goods representer would be useful for various applications. Such applications would include adversarial and non-adversarial scenarios.
[0092] Some non-adversarial applications would include search problems in signal databases, signal monitoring in non-adversarial media. In non-adversarial applications, applying our approach on the whole image would produce favorable results. Yet another application of our algorithm would be several certification applications: In order to compactly describe distinguishing features (face pictures, iris pictures, fingerprints, etc.) of human beings, an application could use their hash values, where the hash values are produced via the exemplary goods representer.

[0093] Exemplary Computing System and Environment [0094] Fig. 3 illustrates an example of a suitable computing environment 300 within which an exemplary goods representer, as described herein, may be implemented (either fully or partially). The computing environment 300 may be utilized in the computer and network architectures described herein.
[0095] The exemplary computing environment 300 is only one example of a computing environment and is not intended to suggest any limitation as to the scope of use or functionality of the computer and network architectures.
Neither should the computing environment 300 be interpreted as having any dependency or requirement relating to any one or combination of components illustrated in the exemplary computing environment 300.
[0096] The exemplary goods representer may be implemented with numerous other general purpose or special purpose computing system environments or configurations. Examples of well known computing systems, environments, and/or configurations that may be suitable for use include, but are not limited to, personal computers, server computers, thin clients, thick clients, hand-held or laptop devices, multiprocessor systems, microprocessor-based systems, set top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments that include any of the above systems or devices, and the like.
[0097] The exemplary goods representer may be described in the general context of processor-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. The exemplary goods representer may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including memory storage devices.
[0098] The computing environment 300 includes a general-purpose computing device in the form of a computer 302. The components of computer 302 may include, but are not limited to, one or more processors or processing units 304, a system memory 306, and a system bus 308 that couples various system components, including the processor 304, to the system memory 306.
[0099] The system bus 308 represents one or more of any of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, and a processor or local bus using any of a variety of bus architectures. By way of example, such architectures can include a CardBus, Personal Computer Memory Card International Association (PCMCIA), Accelerated Graphics Port (AGP), Small Computer System Interface (SCSI), Universal Serial Bus (USB), IEEE 1394, a Video Electronics Standards Association (VESA) local bus, and a Peripheral Component Interconnects (PCI) bus, also known as a Mezzanine bus.
[00100] Computer 302 typically includes a variety of processor-readable media. Such media may be any available media that is accessible by computer 302 and includes both volatile and non-volatile media, removable and non-removable media.
1001011 The system memory 306 includes processor-readable media in the form of volatile memory, such as random access memory (RAM) 310, and/or non-volatile memory, such as read only memory (ROM) 312. A basic input/output system (BIOS) 314, containing the basic routines that help to transfer information between elements within computer 302, such as during start-up, is stored in ROM
312. RAM 310 typically contains data and/or program modules that are immediately accessible to and/or presently operated on by the processing unit 304.
[00102] Computer 302 may also include other removable/non-removable, volatile/non-volatile computer storage media. By way of example, Fig. 3 illustrates a hard disk drive 316 for reading from and writing to a non-removable, non-volatile magnetic media (not shown), a magnetic disk drive 318 for reading from and writing to a removable, non-volatile magnetic disk 320 (e.g., a "floppy disk"), and an optical disk drive 322 for reading from and/or writing to a removable, non-volatile optical disk 324 such as a CD-ROM, DVD-ROM, or other optical media. The hard disk drive 316, magnetic disk drive 318, and optical disk drive 322 are each connected to the system bus 308 by one or more data media interfaces 326. Alternatively, the hard disk drive 316, magnetic disk drive 318, and optical disk drive 322 may be connected to the system bus 308 by one or more interfaces (not shown).
[00103] The disk drives and their associated processor-readable media provide non-volatile storage of computer readable instructions, data structures, program modules, and other data for computer 302. Although the example illustrates a hard disk 316, a removable magnetic disk 320, and a removable optical disk 324, it is to be appreciated that other types of processor-readable media, which may store data that is accessible by a computer, such as magnetic cassettes or other magnetic storage devices, flash memory cards, CD-ROM, digital versatile disks (DVD) or other optical storage, random access memories (RAM), read only memories (ROM), electrically erasable programmable read-only memory (EEPROM), and the like, may also be utilized to implement the exemplary computing system and environment.
[00104] Any number of program modules may be stored on the hard disk 316 magnetic disk 320, optical disk 324, ROM 312, and/or RAM 310, including by way of example, an operating system 326, one or more application programs 328, other program modules 330, and program data 332.
[00105] A user may enter commands and information into computer 302 via input devices such as a keyboard 334 and a pointing device 336 (e.g., a "mouse").
Other input devices 338 (not shown specifically) may include a microphone, joystick, game pad, satellite dish, serial port, scanner, and/or the like.
These and other input devices are connected to the processing unit 304 via input/output interfaces 340 that are coupled to the system bus 308, but may be connected by other interface and bus structures, such as a parallel port, game port, or a universal serial bus (USB).
[00106] A monitor 342 or other type of display device may also be connected to the system bus 308 via an interface, such as a video adapter 344. In addition to the monitor 342, other output peripheral devices may include components, such as speakers (not shown) and a printer 346, which may be connected to computer 302 via the input/output interfaces 340.
[00107] Computer 302 may operate in a networked environment using logical connections to one or more remote computers, such as a remote computing device 348. By way of example, the remote computing device 348 may be a personal computer, portable computer, a server, a router, a network computer, a peer device or other common network node, and the like. The remote computing device 348 is illustrated as a portable computer that may include many or all of the elements and features described herein, relative to computer 302.
[00108] Logical connections between computer 302 and the remote computer 348 are depicted as a local area network (LAN) 350 and a general wide area network (WAN) 352. Such networking environments are commonplace in offices, enterprise-wide computer networks, intranets, and the Internet. Such networking environments may be wired or wireless.
[00109] When implemented in a LAN networking environment, the computer 302 is connected to a local network 350 via a network interface or adapter 354. When implemented in a WAN networking environment, the computer 302 typically includes a modem 356 or other means for establishing communications over the wide network 352. The modem 356, which may be internal or external to computer 302, may be connected to the system bus 308 via the input/output interfaces 340 or other appropriate mechanisms. It is to be appreciated that the illustrated network connections are exemplary and that other means of establishing communication link(s) between the computers 302 and 348 may be employed.
[00110] In a networked environment, such as that illustrated with computing environment 300, program modules depicted relative to the computer 302, or portions thereof, may be stored in a remote memory storage device. By way of example, remote application programs 358 reside on a memory device of remote computer 348. For purposes of illustration, application programs and other executable program components, such as the operating system, are illustrated herein as discrete blocks, although it is recognized that such programs and components reside at various times in different storage components of the computing device 302, and are executed by the data processor(s) of the computer.

Processor-executable Instructions [00111] An implementation of an exemplary goods representer may be described in the general context of processor-executable instructions, such as program modules, executed by one or more computers or other devices.
Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. Typically, the functionality of the program modules may be combined or distributed as desired in various embodiments.

Exemplary Operating Environment [001121 Fig. 3 illustrates an example of a suitable operating environment 300 in which an exemplary goods representer may be implemented. Specifically, the exemplary goods representer(s) described herein may be implemented (wholly or in part) by any program modules 328-330 and/or operating system 326 in Fig. 3 or a portion thereof [00113] The operating environment is only an example of a suitable operating environment and is not intended to suggest any limitation as to the scope or use of functionality of the exemplary goods representer(s) described herein.
Other well known computing systems, environments, and/or configurations that are suitable for use include, but are not limited to, personal computers (PCs), server computers, hand-held or laptop devices, multiprocessor systems, microprocessor-based systems, programmable consumer electronics, wireless phones and equipments, general- and special-purpose appliances, application-specific integrated circuits (ASICs), network PCs, minicomputers, mainframe computers, distributed computing environments that include any of the above systems or devices, and the like.

Processor-readable Media [00114] An implementation of an exemplary goods representer may be stored on or transmitted across some form of processor-readable media. Processor-readable media may be any available media that may be accessed by a computer.

By way of example, processor-readable media may comprise, but is not limited to, "computer storage media" and "communications media."
[00115] "Computer storage media" include volatile and non-volatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules, or other data. Computer storage media includes, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which may be used to store the desired information and which may be accessed by a computer.
[00116] "Communication media" typically embodies processor-readable instructions, data structures, program modules, or other data in a modulated data signal, such as carrier wave or other transport mechanism. Communication media also includes any information delivery media.
[00117] The term "modulated data signal" means a signal that has one or more of its characteristics set or changed in such a manner as to encode information in the signal. By way of example, communication media may comprise, but is not limited to, wired media such as a wired network or direct-wired connection, and wireless media such as acoustic, RF, infrared, and other wireless media. Combinations of any of the above are also included within the scope of processor-readable media.

Conclusion [00118] Although the invention has been described in language specific to structural features and/or methodological steps, it is to be understood that the invention defined in the appended claims is not necessarily limited to the specific features or steps described. Rather, the specific features and steps are disclosed as preferred forms of implementing the claimed invention.

Claims (34)

CLAIMS:
1. A method for producing a new representation of a digital good in a new defined representation domain, the new defined representation domain based upon matrix invariances, the method comprising:
obtaining a digital goods;
selecting, by a partitioner in a computing device, a plurality of pseudo-randomly sized and pseudo-randomly positioned regions from the digital goods;
extracting robust features from the plurality of pseudo-randomly sized and pseudo-randomly positioned regions, wherein the features are based upon singular value decomposition (SVD), discrete cosine transform, or discrete wavelet transform and further wherein the features are within the new defined representation domain;
producing a first output comprising the calculated statistics of one or more of the plurality of pseudo-randomly sized and pseudo-randomly positioned regions, wherein the statistics of the one or more of the plurality of pseudo-randomly sized and pseudo-randomly positioned regions are representatives of respective one or more of the plurality of pseudo-randomly sized and pseudo-randomly positioned regions and are calculated based upon matrix invariances;
constructing a secondary representation of the digital goods by using a pseudo-random combination of the calculated statistics;
forming multiple second regions from the secondary representation;
generating a new set of feature vectors from each second region via a SVD transformation; and producing a second output.
2. A method as recited in claim 1, wherein the digital goods is selected from a group consisting of a digital image, a digital audio clip, a digital video, a database, and a software image.
3. A non-transitory method as recited in claim 1, wherein at least some of the plurality of pseudo-randomly sized and pseudo-randomly positioned regions overlap.
4. A method as recited in claim 1, wherein the plurality of pseudo-randomly sized and pseudo-randomly positioned regions are not allowed to overlap with one another.
5. A system comprising:
a processor;
one or more memories having stored therein processor-executable modules, the modules comprising:
an obtainer configured to obtain a digital goods;
a partitioner configured to pseudo-randomly partition the digital goods into a plurality of regions based at least upon a secret key;
a calculator configured to calculate singular vectors for one or more of the plurality of regions via singular value decompositions, wherein the singular vectors remain invariant with high probability for disturbances to the digital goods in a probability space defined by the secret key.
6. A system as recited in claim 5, wherein at least some of the plurality of regions overlap.
7. A system as recited in claim 5, wherein the digital goods is selected from a group consisting of a digital image, a digital audio clip, a digital video, a database, and a software image.
8. A system as recited in claim 5, wherein the plurality of regions comprise a plurality of pseudo-sized and pseudo-positioned regions.
9. A non-transitory computer storage medium having processor-executable instructions that, when executed by a processor, performs a method comprising: obtaining a digital goods; pseudo-randomly partitioning the digital goods into a plurality of regions; generating singular vectors for one or more of the plurality of regions via singular value decompositions, wherein the singular vectors remain invariant with high probability for disturbances to the digital goods.
10. A non-transitory medium as recited in claim 9, wherein the method further comprises extracting robust pseudo-random features of the digital goods, wherein the features are within a defined representation domain.
11. A non-transitory medium as recited in claim 9, wherein the digital goods is selected from a group consisting of a digital image, a digital audio clip, a digital video, a database, and a software image.
12. A computing device comprising: an output, which is configured to produce output that is audio, visual, or both; and a non-transitory medium as recited in claim 1.
13. A non-transitory computer storage medium having processor-executable instructions that, when executed by a processor, performs a method facilitating protection of digital goods, the method comprising: obtaining a digital goods; partitioning the digital goods into a plurality of pseudo-randomly sized and pseudo-randomly positioned regions, after the partitioning the digital goods, calculating statistics of one or more of the plurality of pseudo-randomly sized and pseudo-randomly positioned regions, wherein the statistics of the one or more of the plurality of pseudo-randomly sized and pseudo-randomly positioned regions are representatives of respective one or more of the plurality of pseudo-randomly sized and pseudo-randomly positioned regions and are calculated based upon matrix invariances; producing output comprising the calculated statistics of the one or more of the plurality of pseudo-randomly sized and pseudo-randomly positioned regions.
14. A non-transitory medium as recited in claim 13, wherein at least some of the plurality of pseudo-randomly sized and pseudo-randomly positioned regions overlap.
15. A non-transitory medium as recited in claim 13, wherein the matrix invariances include singular value decomposition (SVD).
16. A non-transitory medium as recited in claim 13, wherein the digital goods is selected from a group consisting of a digital image, a digital audio clip, a digital video, a database, and a software image.
17. A non-transitory medium as recited in claim 1, wherein pseudo-randomly selecting a plurality of regions comprises selecting a plurality of pseudo-randomly sized and pseudo-randomly positioned regions.
18. A non-transitory medium as recited in claim 1, wherein at least some of the plurality of regions overlap.
19. A non-transitory medium as recited in claim 1, wherein the plurality of regions are not allowed to overlap with one another.
20. A non-transitory medium as recited in claim 13, wherein prior to the partitioning, the method further comprises normalizing amplitudes of the digital goods.
21. A processor-readable medium having processor-executable instructions that, when executed by a processor, performs a method for generating a representation of digital goods, the method comprising:
obtaining a digital good;
segmenting the good into a plurality of regions;

generating feature vectors from each of the regions of the plurality, wherein the feature vectors are calculated based upon matrix invariant singular value decomposition, SVD; and producing output using a combination of the calculated feature vectors, the output forming a hash vector for the digital good.
22. A medium as recited in claim 21, wherein at least some of the plurality of regions overlap.
23. A medium as recited in claim 21, wherein the segmenting comprises pseudo-randomly segmenting the good.
24. A medium as recited in claim 21, wherein the digital good is selected from a group consisting of a digital image, a digital audio clip, a digital video, a database, and a software image.
25. A computing device comprising:
an output, which is configured to produce output that is audio, visual, or both; and a non-transitory medium as recited in claim 21.
26. A computer comprising one or more processor-readable media as recited in claim 21.
27. A method for generating a representation of digital goods, the method comprising:
obtaining a digital good;
segmenting the good into a plurality of regions;

extracting robust feature vectors from each of the regions of the plurality, wherein the feature vectors are calculated based upon matrix invariant singular value decomposition, SVD; and producing output using a combination of the calculated feature vectors, the output forming a hash vector for the digital good.
28. A method as recited in claim 27, wherein at least some of the plurality of regions overlap.
29. A method as recited in claim 27, wherein the segmenting comprises pseudo-randomly segmenting the good.
30. A method as recited in claim 27, wherein the digital good is selected from a group consisting of a digital image, a digital audio clip, a digital video, a database, and a software image.
31. A system comprising:
a processor;
one or more memories having stored therein processor-executable modules, the modules comprising:
an obtainer configured to obtain a digital good;
a partitioner configured to segment the good into a plurality of regions;
and a calculator configured to generate feature vectors from each of the regions of the plurality, wherein the feature vectors are calculated based upon matrix invariant singular value decomposition, SVD; and an output device configured to produce output using a combination of the calculated feature vectors, the output forming a hash vector for the digital good.38
32. A system as recited in claim 31, wherein at least some of the plurality of regions overlap.
33. A system as recited in claim 31, wherein the partitioner is further configured to pseudo-randomly segment the good.
34. A system as recited in claim 31, wherein the digital good is selected from a group consisting of a digital image, a digital audio clip, a digital video, a database, and a software image.
CA2487151A 2004-01-06 2004-11-08 Digital goods representation based upon matrix invariances Expired - Fee Related CA2487151C (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US10/752,268 2004-01-06
US10/752,268 US7831832B2 (en) 2004-01-06 2004-01-06 Digital goods representation based upon matrix invariances

Publications (2)

Publication Number Publication Date
CA2487151A1 CA2487151A1 (en) 2005-07-06
CA2487151C true CA2487151C (en) 2013-05-21

Family

ID=34592560

Family Applications (1)

Application Number Title Priority Date Filing Date
CA2487151A Expired - Fee Related CA2487151C (en) 2004-01-06 2004-11-08 Digital goods representation based upon matrix invariances

Country Status (12)

Country Link
US (1) US7831832B2 (en)
EP (1) EP1553476B1 (en)
JP (1) JP4812291B2 (en)
KR (1) KR101150029B1 (en)
CN (1) CN1638328B (en)
AT (1) ATE425485T1 (en)
AU (1) AU2004237806B2 (en)
BR (1) BRPI0405021A (en)
CA (1) CA2487151C (en)
DE (1) DE602004019876D1 (en)
MX (1) MXPA04012227A (en)
RU (1) RU2387006C2 (en)

Families Citing this family (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8140849B2 (en) * 2004-07-02 2012-03-20 Microsoft Corporation Security for network coding file distribution
CN101855635B (en) * 2007-10-05 2013-02-27 杜比实验室特许公司 Media fingerprints that reliably correspond to media content
US8542869B2 (en) * 2010-06-02 2013-09-24 Dolby Laboratories Licensing Corporation Projection based hashing that balances robustness and sensitivity of media fingerprints
US8776250B2 (en) * 2011-07-08 2014-07-08 Research Foundation Of The City University Of New York Method of comparing private data without revealing the data
EP2549389A1 (en) * 2011-07-20 2013-01-23 Axel Springer Digital TV Guide GmbH Easy 2D navigation in a video database
CN102982804B (en) 2011-09-02 2017-05-03 杜比实验室特许公司 Method and system of voice frequency classification
US9628805B2 (en) * 2014-05-20 2017-04-18 AVAST Software s.r.o. Tunable multi-part perceptual image hashing
CN112668426B (en) * 2020-12-19 2021-11-16 中国民用航空飞行学院 Fire disaster image color cast quantization method based on three color modes

Family Cites Families (127)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DK130099B (en) * 1968-03-16 1974-12-16 Danfoss As Control circuit for an AC motor.
US4773039A (en) * 1985-11-19 1988-09-20 International Business Machines Corporation Information processing system for compaction and replacement of phrases
US5210820A (en) * 1990-05-02 1993-05-11 Broadcast Data Systems Limited Partnership Signal recognition system and method
US5490516A (en) * 1990-12-14 1996-02-13 Hutson; William H. Method and system to enhance medical signals for real-time analysis and high-resolution display
US5093869A (en) * 1990-12-26 1992-03-03 Hughes Aircraft Company Pattern recognition apparatus utilizing area linking and region growth techniques
EP0514688A2 (en) * 1991-05-21 1992-11-25 International Business Machines Corporation Generalized shape autocorrelation for shape acquisition and recognition
US5425081A (en) * 1992-01-22 1995-06-13 Alphanet Telecom Inc. Facsimile arrangement
US5721788A (en) 1992-07-31 1998-02-24 Corbis Corporation Method and system for digital image signatures
US5535020A (en) * 1992-10-15 1996-07-09 Digital Equipment Corporation Void and cluster apparatus and method for generating dither templates
JPH0773190A (en) * 1993-04-29 1995-03-17 Matsushita Electric Ind Co Ltd Pictograph naming for pen base computer system
US7171016B1 (en) 1993-11-18 2007-01-30 Digimarc Corporation Method for monitoring internet dissemination of image, video and/or audio files
US5862260A (en) * 1993-11-18 1999-01-19 Digimarc Corporation Methods for surveying dissemination of proprietary empirical data
US6516079B1 (en) * 2000-02-14 2003-02-04 Digimarc Corporation Digital watermark screening and detecting strategies
US5875264A (en) * 1993-12-03 1999-02-23 Kaman Sciences Corporation Pixel hashing image recognition system
US5465353A (en) * 1994-04-01 1995-11-07 Ricoh Company, Ltd. Image matching and retrieval by multi-access redundant hashing
US5734432A (en) * 1994-07-15 1998-03-31 Lucent Technologies, Inc. Method of incorporating a variable rate auxiliary data stream with a variable rate primary data stream
EP0709766A1 (en) * 1994-10-29 1996-05-01 International Business Machines Corporation Method for the transmission of line-oriented data sets
JPH08186817A (en) * 1994-12-28 1996-07-16 Sony Corp Moving image compressor and its method
US7007166B1 (en) * 1994-12-28 2006-02-28 Wistaria Trading, Inc. Method and system for digital watermarking
US6738495B2 (en) 1995-05-08 2004-05-18 Digimarc Corporation Watermarking enhanced to withstand anticipated corruptions
US6590996B1 (en) * 2000-02-14 2003-07-08 Digimarc Corporation Color adaptive watermarking
US5774588A (en) * 1995-06-07 1998-06-30 United Parcel Service Of America, Inc. Method and system for comparing strings with entries of a lexicon
US5613004A (en) * 1995-06-07 1997-03-18 The Dice Company Steganographic method and device
US5664016A (en) * 1995-06-27 1997-09-02 Northern Telecom Limited Method of building fast MACS from hash functions
US5802518A (en) * 1996-06-04 1998-09-01 Multex Systems, Inc. Information delivery system and method
US5835099A (en) * 1996-06-26 1998-11-10 Xerox Corporation Representing a region of a color image using a space-color separable model
US5778070A (en) * 1996-06-28 1998-07-07 Intel Corporation Method and apparatus for protecting flash memory
US5889868A (en) * 1996-07-02 1999-03-30 The Dice Company Optimization methods for the insertion, protection, and detection of digital watermarks in digitized data
US5918223A (en) * 1996-07-22 1999-06-29 Muscle Fish Method and article of manufacture for content-based analysis, storage, retrieval, and segmentation of audio information
US5915038A (en) * 1996-08-26 1999-06-22 Philips Electronics North America Corporation Using index keys extracted from JPEG-compressed images for image retrieval
WO1998011492A1 (en) * 1996-09-13 1998-03-19 Purdue Research Foundation Authentication of signals using watermarks
US5987159A (en) * 1996-09-24 1999-11-16 Cognex Corporation System or method for detecting defect within a semi-opaque enclosure
US6075875A (en) * 1996-09-30 2000-06-13 Microsoft Corporation Segmentation of image features using hierarchical analysis of multi-valued image data and weighted averaging of segmentation results
US5983351A (en) 1996-10-16 1999-11-09 Intellectual Protocols, L.L.C. Web site copyright registration system and method
US5899999A (en) * 1996-10-16 1999-05-04 Microsoft Corporation Iterative convolution filter particularly suited for use in an image classification and retrieval system
JP3560441B2 (en) * 1997-04-07 2004-09-02 日本アイ・ビー・エム株式会社 Multiple frame data hiding method and detection method
US6081893A (en) * 1997-05-28 2000-06-27 Symantec Corporation System for supporting secured log-in of multiple users into a plurality of computers using combined presentation of memorized password and transportable passport record
US6249616B1 (en) * 1997-05-30 2001-06-19 Enroute, Inc Combining digital images based on three-dimensional relationships between source image data sets
US6131162A (en) * 1997-06-05 2000-10-10 Hitachi Ltd. Digital data authentication method
US5953451A (en) * 1997-06-19 1999-09-14 Xerox Corporation Method of indexing words in handwritten document images using image hash tables
GB9712799D0 (en) 1997-06-19 1997-08-20 Int Computers Ltd Initial program load
US6702417B2 (en) 1997-07-12 2004-03-09 Silverbrook Research Pty Ltd Printing cartridge with capacitive sensor identification
JP4456185B2 (en) 1997-08-29 2010-04-28 富士通株式会社 Visible watermarked video recording medium with copy protection function and its creation / detection and recording / playback device
JP3570236B2 (en) 1997-09-03 2004-09-29 株式会社日立製作所 Image processing method and storage medium storing the program
US5974150A (en) 1997-09-30 1999-10-26 Tracer Detection Technology Corp. System and method for authentication of goods
US6377965B1 (en) * 1997-11-07 2002-04-23 Microsoft Corporation Automatic word completion system for partially entered data
JPH11196262A (en) 1997-11-07 1999-07-21 Matsushita Electric Ind Co Ltd Digital information imbedding extracting device/method, and medium recording program to execute the method
DE19752331C1 (en) * 1997-11-26 1999-09-30 Aesculap Ag & Co Kg Magazine for a surgical clip applier
US6330672B1 (en) 1997-12-03 2001-12-11 At&T Corp. Method and apparatus for watermarking digital bitstreams
US6101602A (en) * 1997-12-08 2000-08-08 The United States Of America As Represented By The Secretary Of The Air Force Digital watermarking by adding random, smooth patterns
JP3986150B2 (en) * 1998-01-27 2007-10-03 興和株式会社 Digital watermarking to one-dimensional data
US6513118B1 (en) * 1998-01-27 2003-01-28 Canon Kabushiki Kaisha Electronic watermarking method, electronic information distribution system, image filing apparatus and storage medium therefor
KR100302366B1 (en) 1998-02-14 2001-11-30 이계철 Apparatus and method for searching layout base image
US7234640B2 (en) * 1998-04-17 2007-06-26 Remote Inc. Portable ordering device
US6314192B1 (en) 1998-05-21 2001-11-06 Massachusetts Institute Of Technology System, method, and product for information embedding using an ensemble of non-intersecting embedding generators
JP3809297B2 (en) 1998-05-29 2006-08-16 キヤノン株式会社 Image processing method, apparatus and medium
US6285995B1 (en) * 1998-06-22 2001-09-04 U.S. Philips Corporation Image retrieval system using a query image
US6144958A (en) * 1998-07-15 2000-11-07 Amazon.Com, Inc. System and method for correcting spelling errors in search queries
US6658626B1 (en) 1998-07-31 2003-12-02 The Regents Of The University Of California User interface for displaying document comparison information
JP2000115728A (en) 1998-10-07 2000-04-21 Sony Corp Video signal transmission system, video signal output device, video signal processor and video signal transmission method
US6256409B1 (en) * 1998-10-19 2001-07-03 Sony Corporation Method for determining a correlation between images using multi-element image descriptors
US6363381B1 (en) * 1998-11-03 2002-03-26 Ricoh Co., Ltd. Compressed document matching
JP2000149004A (en) 1998-11-10 2000-05-30 Matsushita Electric Ind Co Ltd Image reader
US7104449B2 (en) * 1998-11-12 2006-09-12 Wenyu Han Method and apparatus for patterning cards, instruments and documents
US6321232B1 (en) 1998-12-18 2001-11-20 Xerox Corporation Method for creating a geometric hash tree in a document processing system
GB2363300B (en) * 1998-12-29 2003-10-01 Kent Ridge Digital Labs Digital audio watermarking using content-adaptive multiple echo hopping
US6442283B1 (en) * 1999-01-11 2002-08-27 Digimarc Corporation Multimedia data embedding
US6532541B1 (en) * 1999-01-22 2003-03-11 The Trustees Of Columbia University In The City Of New York Method and apparatus for image authentication
JP2000332988A (en) 1999-05-19 2000-11-30 Matsushita Electric Ind Co Ltd Device and method for embedding and extracting digital information and medium with program for executing the method recorded thereon
WO2000043910A1 (en) * 1999-01-22 2000-07-27 Kent Ridge Digital Labs Method and apparatus for indexing and retrieving images using visual keywords
US6278385B1 (en) * 1999-02-01 2001-08-21 Yamaha Corporation Vector quantizer and vector quantization method
JP2000243067A (en) 1999-02-19 2000-09-08 Sony Corp System, device and method for audio editing and recording medium
JP3740314B2 (en) 1999-03-11 2006-02-01 キヤノン株式会社 Image processing apparatus and method
US6246777B1 (en) * 1999-03-19 2001-06-12 International Business Machines Corporation Compression-tolerant watermarking scheme for image authentication
KR100333163B1 (en) * 1999-03-29 2002-04-18 최종욱 Digital watermarking method and apparatus
US6331859B1 (en) * 1999-04-06 2001-12-18 Sharp Laboratories Of America, Inc. Video skimming system utilizing the vector rank filter
US6901514B1 (en) * 1999-06-01 2005-05-31 Digital Video Express, L.P. Secure oblivious watermarking using key-dependent mapping functions
JP2000350007A (en) 1999-06-03 2000-12-15 Ricoh Co Ltd Electronic watermarking method, electronic watermark device and recording medium
US6418430B1 (en) * 1999-06-10 2002-07-09 Oracle International Corporation System for efficient content-based retrieval of images
US6782361B1 (en) * 1999-06-18 2004-08-24 Mcgill University Method and apparatus for providing background acoustic noise during a discontinued/reduced rate transmission mode of a voice transmission system
US6768980B1 (en) * 1999-09-03 2004-07-27 Thomas W. Meyer Method of and apparatus for high-bandwidth steganographic embedding of data in a series of digital signals or measurements such as taken from analog data streams or subsampled and/or transformed digital data
US6574348B1 (en) * 1999-09-07 2003-06-03 Microsoft Corporation Technique for watermarking an image and a resulting watermarked image
US6546114B1 (en) * 1999-09-07 2003-04-08 Microsoft Corporation Technique for detecting a watermark in a marked image
US6751343B1 (en) * 1999-09-20 2004-06-15 Ut-Battelle, Llc Method for indexing and retrieving manufacturing-specific digital imagery based on image content
US6671407B1 (en) 1999-10-19 2003-12-30 Microsoft Corporation System and method for hashing digital images
US6606744B1 (en) * 1999-11-22 2003-08-12 Accenture, Llp Providing collaborative installation management in a network-based supply chain environment
US7016540B1 (en) * 1999-11-24 2006-03-21 Nec Corporation Method and system for segmentation, classification, and summarization of video images
US6725372B1 (en) * 1999-12-02 2004-04-20 Verizon Laboratories Inc. Digital watermarking
ATE504063T1 (en) * 1999-12-24 2011-04-15 Ibm METHOD AND SYSTEM FOR DETECTING IDENTICAL DIGITAL DATA
US6769061B1 (en) * 2000-01-19 2004-07-27 Koninklijke Philips Electronics N.V. Invisible encoding of meta-information
US6385329B1 (en) 2000-02-14 2002-05-07 Digimarc Corporation Wavelet domain watermarks
US6584465B1 (en) * 2000-02-25 2003-06-24 Eastman Kodak Company Method and system for search and retrieval of similar patterns
US6701014B1 (en) * 2000-06-14 2004-03-02 International Business Machines Corporation Method and apparatus for matching slides in video
AU2001275887A1 (en) * 2000-07-12 2002-01-21 Cornell Research Foundation Inc. Method and system for analyzing multi-variate data using canonical decomposition
US6990453B2 (en) * 2000-07-31 2006-01-24 Landmark Digital Services Llc System and methods for recognizing sound and music signals in high noise and distortion
RU2193825C2 (en) 2000-08-10 2002-11-27 Открытое акционерное общество "Научно-конструкторское бюро вычислительных систем" Method and device for processing signals to find coordinates of objects displayed as sequence of television images
US20040064416A1 (en) * 2000-10-03 2004-04-01 Ariel Peled Secure distribution of digital content
US6907527B1 (en) * 2000-10-17 2005-06-14 International Business Machines Corporation Cryptography-based low distortion robust data authentication system and method therefor
WO2002051063A1 (en) * 2000-12-21 2002-06-27 Digimarc Corporation Methods, apparatus and programs for generating and utilizing content signatures
US6879703B2 (en) * 2001-01-10 2005-04-12 Trustees Of Columbia University Of The City Of New York Method and apparatus for watermarking images
US6990444B2 (en) * 2001-01-17 2006-01-24 International Business Machines Corporation Methods, systems, and computer program products for securely transforming an audio stream to encoded text
US6658423B1 (en) 2001-01-24 2003-12-02 Google, Inc. Detecting duplicate and near-duplicate files
US7020775B2 (en) 2001-04-24 2006-03-28 Microsoft Corporation Derivation and quantization of robust non-local characteristics for blind watermarking
US6975743B2 (en) 2001-04-24 2005-12-13 Microsoft Corporation Robust and stealthy video watermarking into regions of successive frames
US6996273B2 (en) * 2001-04-24 2006-02-07 Microsoft Corporation Robust recognizer of perceptually similar content
US6973574B2 (en) * 2001-04-24 2005-12-06 Microsoft Corp. Recognizer of audio-content in digital signals
US6654740B2 (en) * 2001-05-08 2003-11-25 Sunflare Co., Ltd. Probabilistic information retrieval based on differential latent semantic space
US6915009B2 (en) * 2001-09-07 2005-07-05 Fuji Xerox Co., Ltd. Systems and methods for the automatic segmentation and clustering of ordered information
US7398395B2 (en) * 2001-09-20 2008-07-08 Koninklijke Philips Electronics N.V. Using multiple watermarks to protect content material
JP4035383B2 (en) * 2001-10-22 2008-01-23 株式会社リコー Digital watermark code generation apparatus and code generation method, digital watermark decoding apparatus and decoding method, digital watermark code generation and decoding program, and recording medium recording the same
JP3953295B2 (en) * 2001-10-23 2007-08-08 インターナショナル・ビジネス・マシーンズ・コーポレーション Information search system, information search method, program for executing information search, and recording medium on which program for executing information search is recorded
US7006658B2 (en) * 2001-12-20 2006-02-28 Koninklijke Philips Electronics N.V. Varying segment sizes to increase security
US7062419B2 (en) * 2001-12-21 2006-06-13 Intel Corporation Surface light field decomposition using non-negative factorization
US9031128B2 (en) * 2001-12-31 2015-05-12 Stmicroelectronics Asia Pacific Pte Ltd. Video encoding
US7142675B2 (en) 2002-02-12 2006-11-28 City University Of Hong Kong Sequence generator and method of generating a pseudo random sequence
US20030169289A1 (en) * 2002-03-08 2003-09-11 Holt Duane Anthony Dynamic software control interface and method
US6919896B2 (en) 2002-03-11 2005-07-19 Sony Computer Entertainment Inc. System and method of optimizing graphics processing
US7133538B2 (en) * 2002-04-10 2006-11-07 National Instruments Corporation Pattern matching utilizing discrete curve matching with multiple mapping operators
US7139432B2 (en) * 2002-04-10 2006-11-21 National Instruments Corporation Image pattern matching utilizing discrete curve matching with a mapping operator
US7327887B2 (en) * 2002-04-10 2008-02-05 National Instruments Corporation Increasing accuracy of discrete curve transform estimates for curve matching
US6864897B2 (en) * 2002-04-12 2005-03-08 Mitsubishi Electric Research Labs, Inc. Analysis, synthesis and control of data signals with temporal textures using a linear dynamic system
KR100754721B1 (en) 2002-04-26 2007-09-03 삼성전자주식회사 Apparatus and method for transmitting and receiving multiplexed data in an orthogonal frequency division multiplexing communication system
US7095873B2 (en) * 2002-06-28 2006-08-22 Microsoft Corporation Watermarking via quantization of statistics of overlapping regions
US6999074B2 (en) * 2002-11-22 2006-02-14 Intel Corporation Building image-based models by mapping non-linear optimization to streaming architectures
US20050165690A1 (en) * 2004-01-23 2005-07-28 Microsoft Corporation Watermarking via quantization of rational statistics of regions
US20050163313A1 (en) * 2004-01-23 2005-07-28 Roger Maitland Methods and apparatus for parallel implementations of table look-ups and ciphering
US20070053325A1 (en) * 2005-04-26 2007-03-08 Interdigital Technology Corporation Method and apparatus for securing wireless communications

Also Published As

Publication number Publication date
RU2004135760A (en) 2006-05-20
CA2487151A1 (en) 2005-07-06
US20050149727A1 (en) 2005-07-07
CN1638328A (en) 2005-07-13
KR101150029B1 (en) 2012-05-30
BRPI0405021A (en) 2005-09-20
KR20050072394A (en) 2005-07-11
RU2387006C2 (en) 2010-04-20
ATE425485T1 (en) 2009-03-15
US7831832B2 (en) 2010-11-09
CN1638328B (en) 2011-09-28
AU2004237806A1 (en) 2005-07-21
EP1553476A2 (en) 2005-07-13
EP1553476B1 (en) 2009-03-11
JP4812291B2 (en) 2011-11-09
DE602004019876D1 (en) 2009-04-23
JP2005196744A (en) 2005-07-21
MXPA04012227A (en) 2005-07-07
AU2004237806B2 (en) 2010-02-18
EP1553476A3 (en) 2006-10-04

Similar Documents

Publication Publication Date Title
Hosny et al. Robust color image watermarking using invariant quaternion Legendre-Fourier moments
US7266244B2 (en) Robust recognizer of perceptually similar content
Davarzani et al. Perceptual image hashing using center-symmetric local binary patterns
Bhatnagar et al. Biometrics inspired watermarking based on a fractional dual tree complex wavelet transform
US7577272B2 (en) Digital fingerprinting using synchronization marks and watermarks
US20070076869A1 (en) Digital goods representation based upon matrix invariants using non-negative matrix factorizations
Aparna et al. A blind medical image watermarking for secure E-healthcare application using crypto-watermarking system
Singh et al. Superpixel based robust reversible data hiding scheme exploiting Arnold transform with DCT and CA
US20050165690A1 (en) Watermarking via quantization of rational statistics of regions
CA2487151C (en) Digital goods representation based upon matrix invariances
Agarwal et al. Image watermarking in real oriented wavelet transform domain
Pilania et al. An ROI-based robust video steganography technique using SVD in wavelet domain
Khaldi et al. Deformable model segmentation for range image watermarking
Singh et al. Guest editorial: robust and secure data hiding techniques for telemedicine applications
Rahardi et al. A Blind Robust Image Watermarking on Selected DCT Coefficients for Copyright Protection
Singh et al. Robust watermarking scheme for compressed image through dct exploiting superpixel and arnold transform
Baumy et al. Efficient Forgery Detection Approaches for Digital Color Images.
US7136535B2 (en) Content recognizer via probabilistic mirror distribution
Madhavan et al. Optimisation for video watermarking using ABC algorithm
Sun et al. Digital watermarks for videos based on a locality-sensitive hashing algorithm
Radhakrishnan et al. On the security of the visual hash function
Bhatnagar Robust covert communication using high capacity watermarking
Rajput et al. A robust watermarking scheme via optimization-based image reconstruction technique
Preetha et al. A Wavelet Optimized Video Copy Detection Using Content Fingerprinting
Selvy et al. A novel watermarking of images based on wavelet based contourlet transform energized by biometrics

Legal Events

Date Code Title Description
EEER Examination request
MKLA Lapsed

Effective date: 20210831

MKLA Lapsed

Effective date: 20191108