MAIWALD · Elisenstraße 3 · 80335 München Maiwald Patentanwalts- und

Rechtsanwaltsgesellschaft mbH
Europäisches Patentamt
München
80298 München Düsseldorf

Elisenstraße 3
80335 München
T +49 89 7472660
F +49 89 776424
H www.maiwald.eu

Munich, 8 October 2020 Dr. Simon Quartus Lud

Partner
lud@maiwald.eu
Official Ref.: EP 18 185 240.1 / 3 410 716
Patentee: GE Video Compression, LLC
Opponent: Unified Patents, LLC

Our Ref.: U07933EPOP/SQL

Geschäftsführer
(Board of Directors)

Grounds of Opposition Dr. Regina Neuefeind, LL.M.

Lutz Kietzmann, LL.M.
Dr. Norbert Hansen
Dr. Martin Huenges, LL.M.
Dr. Holger Glas, LL.M.
Dr. Dirk Bühler
Dr. Christian Schäflein
This opposition is filed against grant of European Patent EP 3 410 716 B1 titled Dr. Alexander Schmitz
Angela Zumstein
Dr. Derk Vos
“CODING OF SIGNIFICANCE MAPS AND TRANSFORM COEFFICIENT Dr. Berthold Lux
Dr. Eva Ehlich
Alexander Ortlieb
BLOCKS” (in German: “CODIERUNG VON SIGNIFIKANZMATRIZEN UND Dr. Andreas Ledl
Dr. Eva Dörner
TRANSFORMATIONSKOEFFIZIENTENBLÖCKEN”) in the name and on behalf Dr.-Ing. Sophie Ertl
Dr. Marco Stief, LL.M.
Dr. Ralf Kotitschke
of “the Opponent”, Dr. Christian Pioch

HRB Nr. 111307

Amtsgericht München
Unified Patents, LLC
1875 Connecticut Ave., NW, Floor 10
Washington, D.C. 20009

SQL:SQL
Table of contents

A. Introduction and Formalities ..................................................................3

1. Requests....................................................................................................3
2. The Prior Art ..............................................................................................3

B. European Patent EP 3 410 716 B1..........................................................5

1. Formalities .................................................................................................5
2. Subject matter of the alleged invention......................................................6
3. Technological Background ........................................................................9
3.1 Partitioning........................................................................................10
3.2 Prediction..........................................................................................10
3.3 Transform and Quantization .............................................................10
3.4 Significance Map ..............................................................................11
4. The person skilled in the art.....................................................................14
5. Granted independent claims....................................................................15
5.1 Granted claim 1 ................................................................................15
5.2 Granted claim 3 ................................................................................16

C. Grounds for Opposition ........................................................................18

I. Added subject matter – Articles 100 (c), 123 (2) EPC and Articles 100
(c), 76 EPC.........................................................................................................18
II. Lack of novelty – Articles 100 (a), 52 EPC ..............................................19
1. Lack of novelty over cited document D2..................................................19
1.1 Independent claim 1 .........................................................................19
1.2 Independent claim 3 .........................................................................20
1.3 Independent claims 5 and 6 .............................................................20
1.4 Independent claims 7 and 9 .............................................................20
III. Lack of inventive step – Articles 100 (a), 56 EPC ...............................21
1. Lack of inventive step over document D4 taken alone or in combination
with D5 ............................................................................................................21
1.1 Independent claim 1 .........................................................................21
1.2 Independent claim 3 .........................................................................24
1.3 Independent claims 5 and 6 .............................................................25
1.4 Independent claims 7 and 9 .............................................................25
1.5 Dependent claims.............................................................................26
1.5.1 Claim 2 ......................................................................................26
IV. Sufficiency of Disclosure – Articles 100 (b), 83 EPC.........................27

D. Summary and Conclusion ....................................................................29

2
A. Introduction and Formalities

1 The Opponent hereby opposes European Patent EP 3 410 716 B1

(hereinafter “the Patent”) granted on 08 January 2020 to GE Video
Compression, LLC, Albany, NY 12211 (US) (hereinafter “the Patentee”).

1. Requests

2 It is hereby requested that the Patent be revoked in its entirety under

Article 99 EPC on the grounds of:

(a) Article 100(a) EPC because the subject matter of the Patent lacks
novelty contrary to Articles 52(1), 54 EPC and/or lacks inventive
step contrary to Articles 52(1), 56 EPC; and

(b) Article 100(b) EPC because the subject matter of the Patent lacks
sufficiency of disclosure contrary to Articles 83 EPC; and

(c) Article 100(c) EPC because the subject matter of the Patent
extends beyond the content of the European Application as
originally filed contrary to Article 123(2) EPC.

3 In the event that the Opposition Division is not in a position to revoke the
Patent in its entirety (see request (1) above), Oral Proceedings pursuant
to Article 116 EPC are hereby requested.

2. The Prior Art

4 In these grounds of opposition, the Opponent relies upon the following

documents:

Ref. Publication No.

D1 EP1487113 (“Marpe”)
D2 SALOMON D et al, Data Compression: The
Complete Reference (passage), New York,
NY, US, Springer, (1998), pages 69 - 84,
ISBN 978-0-387-98280-9, XP002270343
(“Salomon”)
D3 ZHANG L ET AL, "Context-based Arithmetic
Coding Reexamined for DCT Video

3
 Compression", PROCEEDINGS OF THE
2007 IEEE INTERNATIONAL SYMPOSIUM
ON CIRCUITS AND SYSTEMS (ICASP
2007), IEEE, Piscataway, NJ, US,
(20070501), ISBN 978-1-4244-0920-4, pages
3147 - 3150, XP031181972 (“Zhang”)
D4 US Pat. Pub. 2008/0219578 to Lee (“Lee”)
D5 US Pat. Pub. 2008/0310745 to Ye et al.(“Ye”)
D6 “The Quadtree and Related Hierarchical Data
Structures”, by Hanan Samet, ACM
Computing Surveys, June 1984
Vol. 16, No. 2, June 1984 (“Computing
Surveys”)
D7 IEEE Transactions on Circuits and Systems
For Video Technology, Context-Based
Adaptive Binary Arithmetic Coding in the
H.264/AVC Video Compression Standard, by
Marpe et al. (2003) (“IEEE CABAC”)
D8 PENNEBAKER W B ET AL, "An overview of
the basic principles of the Q-Coder adaptive
binary arithmetic coder", IBM JOURNAL OF
RESEARCH AND DEVELOPMENT, IBM
Corporation, New York, NY, US, (198811),
vol. 32, no. 6, ISSN 0018-8646, pages 717 –
726 (“Pennebaker”)
D9 Shi et al., Image and Video Compression for
Multimedia Engineering, CRC Press (2000)
(“Shi”)
D10 Very Low Complexity MPEG-2 to H.264
Transcoding Using Machine Learning, by
Fernandez et al. (2006) (“H.264 Transcoding”)
D11 BRADY N ET AL, "Context-based Arithmetic
Encoding of 2D Shape Sequences",
PROCEEDINGS OF THE IEEE
INTERNATIONAL CONFERENCE ON IMAGE
PROCESSING (ICIP 1997), IEEE, Los
Alamitos, CA, US, (19971026), vol. 1,
doi:10.1109/ICIP.1997.647376, ISBN 978-0-
8186-8183-7, pages 29 - 32, XP010254100
[A] 1-9 * abstract * * page 30, column l, line 8 -
column r, line 6 * (“Brady”)
D12 Video Coding Basics Presentation by Wang
(2003) (“Video Coding Basics”)
D13 IEEE Transactions on Circuits and Systems
For Video Technology, Overview of the
H.264/AVC Video Coding Standard, by
Wiegand et al. (2003) (“IEEE H.264”)
D14 A Context Modeling Algorithm And Its
Application In Video Compression by Mrak et
al. (2003) (“Context Modeling”)

4
D15 Adaptive Scanning for H.264/AVC Intra
Coding, by Lee et al. (2006) (“Adaptive
Scanning”)
D16 Introduction to Statistics by David M. Lane
(2007)
D17 Declaration of Dr. Andrew B. Lippman

5 The list of references is annexed hereto as

Annex M1.

6 Each of the documents listed above was published before the earliest
priority date of the Patent, which is 13.04.2010.

7 Therefore, each of cited documents D1 to D16 constitutes prior art under

Article 54(2) EPC, thus relevant for the assessment of both lack of novelty
and inventive step.

8 Since the vast majority of the cited documents was available to the public
prior to the claimed priority date, the issue of entitlement to priority is not
discussed in detail herein below.

9 However, this is no admission that the claimed subject matter would

validly claim the priority.

10 References to the cited prior art documents are to paragraphs or sections

of the publications listed above. These references, however, are
exemplary and not exhaustive.

11 The Opponent herewith reserves the right to refer to other sections of the
prior art documents if necessary later in the proceedings.

12 Cited document D17 is a declaration of Dr. Andrew B. Lippman, which is

submitted herewith as an expert opinion.

13 The declaration was used in the PETITION FOR INTER PARTES

REVIEW, OF U.S. PATENT NO. 9,357,217 Case No. IPR2020-00858.

5
B. European Patent EP 3 410 716 B1

In the following, bibliographic and formal aspects of the Patent, EP 3 410

716 B1, as well as its subject matter are discussed.

1. Formalities

14 The Patent, EP 3 410 716 B1, derives from European Patent Application
Number 18185240.1 (hereinafter “the European Application”) which has
a filing date of 11 April 2011.

15 The European Application was published as EP 3 410 716 A1 on

05.12.2018. The European Application was filed on 26.07.2016 as a
divisional application to the applications EP11713791.9 / EP2559244 and
EP17177435.9 / EP3244612.

16 The Patent claims the priority date of 13.04.2010, which are the filing
dates of Patent Applications PCT/EP2010/054822 and EP 10159766
(hereinafter “the Priority Applications”).

17 In the following, references to the Patent and paragraphs thereof (shown

in square brackets […]) relate to the published B1-specification, unless
specified otherwise.

2. Subject matter of the alleged invention

18 The alleged invention generally concerns systems and methods for to

coding of significance maps indicating positions of significant transform
coefficients within transform coefficient blocks and the coding of such
transform coefficient blocks.

19 Such coding may, for example, be used in picture and video coding, for
example.

20 As previewed above, the purportedly novel aspect of the coding process

described and claimed in the Patent relates to the idea of adapting the
scan order of a plurality of sub-blocks of a transform block in order to take
advantage of the grouping, or concentration, of significant transform
coefficients that frequently occurs.

21 The concept is reflected, for example, in the following exemplary limitation

of Claim 1 of the Patent, emphasis added:

6
 “extracting a significance map indicating positions of significant
transform coefficients within the transform coefficient block, and
then the values of the significant transform coefficients within the
transform coefficient block from a data stream, with, in extracting the
significance map, sequentially extracting first-type syntax
elements from the data stream by context-adaptive entropy
decoding, the first-type syntax elements indicating, for
associated positions within the transform coefficient block as
to whether at the respective position a significant or
insignificant transform coefficient is situated.”

22 For purposes of understanding the alleged novel concept, it is helpful to

picture a concentrated group of significant transform coefficients within a
transform block, as displayed by a generic block below:

23 Thus, the basic concept is to select a scan order that encounters the
significant coefficients early in the scan. The Patent describes the
purported advantages of this adaptive scan order at [0051] et seq. of the
Patent. In particular, the Patent therein in details explains, emphasis
added:

“The advantage of the switching between scan patterns is a

reduced bit rate, which is a result of a smaller number of coded
syntax elements. As an intuitive example and referring to FIG. 6, it
is often the case that significant transform coefficient
values—in particular for large transform blocks—are
concentrated at one of the block borders 270, 272, because the
residual blocks contain mainly horizontal or vertical structures.
With the mostly used zig-zag scan 274, there exists a probability of
about 0.5 that the last diagonal sub-scan of the zig-zag scan in
which the last significant coefficient is encountered starts from the
side at which the significant coefficients are not concentrated. In
that case, a large number of syntax elements for transform

7
 coefficient levels equal to zero have to be coded before the
last non-zero transform coefficient value is reached. This can
be avoided if the diagonal sub-scans are started at the side,
where the significant transform coefficient levels are
concentrated.”

24 A variety of scan orders could be utilized to accomplish this goal of

encountering the concentrated significant coefficients early in the scan,
including zig-zag, reverse zig-zag, diagonal, and so on. Different scan
orders can yield different run lengths that better exploit the concentrated
coefficients, see in this regard for instance D1 at [0014].

25 Figure 11 of the Patent depicts macroblock 256, which consists of four

sub-blocks 322. Each sub-block 322 consists of 16 (4x4) blocks. In this
example, the ’217 patent describes using zig-zag scanning within macro
block 256 and reverse zig-zag scanning within sub-blocks 322:

26

27 This reverse zig-zag scan order for the sub-blocks may be selected, for
example, if there is a predicted concentration of transform coefficients in
the lower right quadrant of each sub-block.

28 It is frequently the case that high-energy transform coefficients typically

reside near the corners of transform blocks.

29 Thus, zig-zag scanning (either forward or backward) became widely

utilized as it represented an effective way to group non-zero coefficients

8
 near the beginning of the scan order, see in this regard for instance cited
document D7 at page 10, “zig-zag scanning pattern”, left column.

30 However, significant transform coefficients do not always reside in the

same region, so zig-zag scanning does not always achieve the most
effective scanning of concentrated coefficients. Thus, the claims of the
Patent recite adaptive selection of the scan order, with a diagonal scan
order being one of the possible options.

31 Examples of various diagonal scan orders can be seen in Fig. 8 of the

Patent:

32 By adaptively determining the scan order, including by considering

diagonal scan orders, based on coefficient concentrations in nearby
previously decoded blocks, the Patent attempts to take advantage of
concentrations of significant coefficients that may not be captured by
using only a fixed scan order, such as zig-zag scan order. However, this
adaptive selection of scan orders, including consideration of diagonal
scan orders, was already widely known by the time of the Patent.

3. Technological Background

33 Digital video is formed from a sequence of video frames that include

picture element (or pixel) data. During playback, the frames are
successively displayed at a certain frame rate, rendering the video for
display.

34 Each frame is an array of pixels organized in rows and columns to form

the image represented by the frame, reflecting characteristics of objects
represented in a scene of a video. Video files can be very large due to
large amounts of image data associated with each frame.

9
 35 Therefore, video coding techniques are used to compress (or encode)
video files for more manageable storage and transmission. Such
compression is achieved by removing redundancy in and between
frames, thereby reducing the amount of information necessary to transmit
and reconstruct each frame.

36 Transform coefficient blocks and significance maps are used to convey

where these redundancies lie and to take advantage of the same.

37 The coding process generally consists of four phases/processes—(1)

partitioning; (2) prediction; (3) transform and quantization; and (4) entropy
encoding—all of which were well known and had been practiced long
before the effective filing date of the Patent, see in this regard for instance
cited document D7, Chapter III. DETAILED DESCRIPTION OF CABAC.

3.1 Partitioning

38 Partitioning is the process of dividing a frame into blocks of pixels. These

blocks can vary in size, as illustrated below:

H.264 Transcoding (D10) at 932.

3.2 Prediction

39 The encoder performs prediction of each partitioned block of the frame

based on coded images from previous frames (inter prediction) or
previously coded regions of the same frame (intra prediction), See D7,
Section titled “Coding of Prediction Modes”, page 9.

40 While there are multiple styles of prediction, the general concept is to

search previously coded blocks to locate the block with the most similarity
to the current block and to use that as a reference block, or prediction, for
the current block.

41 The predicted samples are then subtracted from the samples of the
current block, resulting in a difference signal (i.e., a residual). This

10
 residual only contains information about the differences between the
images in the current block and the predictive reference block.

42 The residual contains significantly less information than the block being
encoded, reducing the amount of data to be compressed and transmitted.
Alternatively, regions that cannot be predicted with good efficiency are
coded using intraframe prediction.

3.3 Transform and Quantization

43 The block of residual samples is transformed into a set of values (called

transform coefficients) through a transform equation, a common example
of which is Discrete Cosine Transform (DCT).

44 The resulting block of transform coefficients can then be quantized

wherein each coefficient is divided by an integer value to remove
insignificant transform coefficients.

45 Quantization reduces the precision of the transform coefficients

according to a quantization parameter (QP).

46 Typically, the result is a block in which most or all of the coefficients are
zero, with a few non-zero coefficients, as displayed by a generic block
below:

47 Setting QP to a high value means that more coefficients are set to zero,
resulting in high compression at the expense of poor quality of the
decoded image.

11
 48 Setting QP to a low value means that more non-zero coefficients remain
after quantization, resulting in better decoded image quality but lower
compression.

49 The result of the transform and quantization process is generally referred

to as a transform block or unit or single block of transform coefficient, see
Fig. 5 of D7. See also cited document D12, 13 and 14 in this context.

3.4 Significance Map

50 A significance map is a matrix of the same size as the transform unit, with
each entry being a coefficient flag corresponding to a quantized transform
coefficient in the transform unit, see page 11, left column of D7.

51 A significance map is a representation of positions of significant transform

coefficients within a transform coefficient block, which allows the decoder
to more efficiently know where meaningful data exists without having to
fully process blocks with insignificant transform coefficients.

52 Transform coefficients are “significant” when their value is high enough

such that omitting those coefficients will result in a noticeable reduction in
picture quality.

53 Conversely, transform coefficients are insignificant when their value is low

enough such that omitting those coefficients altogether in a coding
process will not result in a noticeable reduction in picture quality.

12
54 Each coefficient flag is either a “1” if it corresponds to a significant
coefficient or a “0” if it corresponds to an insignificant coefficient. A
significance map is illustrated below (from the “Lee” reference discussed
herein, e.g. Fig. 3A and 3B of D4):

55 During entropy coding, both the value of the transform coefficients that
result from the above-noted prediction, transform, and quantization
process as well as their significance map are encoded to generate the
bitstream.

56 The bitstream thus includes a string of 1s and 0s (indicating the locations

of the significant coefficients) as well as information about the magnitudes
and signs of only the significant coefficients. Because only significant
coefficients are meaningful, this limits the amount of data that must be
placed into the bitstream, enhancing overall compressions, see in this
regard also cited document D4 at paragraphs [0040] to [0046].

57 Typically, the significance map is translated from a two-dimensional map

into a one-dimensional vector (i.e., a bitstream) by scanning a transform
unit in a particular order (e.g., diagonal, vertical, horizontal, and zigzag
patterns) to capture the information about the locations, magnitudes and
signs of the significant coefficients, see in this regard also cited document
D4 at ¶ 46.

58 The same process can be reversed on the receiving, or decoding, side.

Some scanning methods are illustrated below:

13
 IEEE H.264 at 568 (D13).

59 As will be discussed further below in connection with the Patent,

significant transform coefficients are often bunched together (i.e., a
current block frequently differs from a predictive reference block in one
primary area of the image) in the significance map.

60 Further efficiencies can be gained by adapting the scan patterns to

similarly bunch the significant transform coefficients in the bitstream such
that the remaining insignificant transform coefficients can quickly be set
aside.

61 On the receiving end, the bitstream is decoded by a decoder to

reconstruct the significance map, as well as the associated transform unit,
and this decoded transform unit is then used in connection with a
previously decoded reference frame to reproduce the video frame.

62 Context-adaptive coding, e.g., CABAC, may then be applied to achieve

even more compression of the resulting bitstream, see in this regard cited
document D7.

63 A context is a probability model indicating how often certain binary values

appear in the bitstream.

64 According to the selected context, values that appear more frequently are
encoded using the smallest code words to represent the information, and
values that appear less frequently can be encoded with larger code words
to represent the less frequently appearing information.

14
 65 By using a context (or probability model) as a guide, the total length of
code words is reduced, and compression is further enhanced. As
suggested by the name, CABAC adapts the probability model used based
on information from nearby elements to optimize the probability estimate.

4. The person skilled in the art

66 The person skilled in the art would have been a person having ordinary
skill in the art and would be a person having, as of April 13, 2010 at least
an undergraduate degree in computer science, computer engineering,
electrical engineering, or a similar technical field.

67 The person skilled in the art would have further gained a working
knowledge of video coding techniques; and two or more years of
experience (or with a graduate degree in the above-stated fields, one or
more years of experience) in analysis, design, or development related to
video encoding, with additional education substituting for experience and
vice-versa.

5. Granted independent claims

68 The Patent comprises an independent device claim 1 related to “a

decoder for decoding a data stream” and an independent device claim 3
related to “an encoder for encoding an array of information samples” and
an independent method claim 5 related to “a method for decoding a data
stream” and an independent method claim 6 related to “a method for
encoding an array of information samples into a data stream in units of
blocks”.

69 Further, the Patent comprises an independent claim 7 related to “a data

stream” and an independent memory product claim 9 related to “a
computer readable digital storage medium” referring back to the methods
claims.

5.1 Granted claim 1

70 For ease of reference, independent device claim 1 of the Patent is shown

below, divided into separate integers in the form of a feature analysis.

1.1 Decoder for decoding a data stream (30) to reconstruct a

reconstructed version of a sample array (20), configured to
extract from the data stream a transform coefficient block for

15
 each of blocks (40) of the sample array, and reconstruct the
reconstructed version by performing an inverse transform on
each of the transform coefficient blocks,

1.2 extract, by entropy decoding, information on the transform

coefficient blocks from the data stream by decoding a
significance map indicating positions of significant transform
coefficients within a transform coefficient block from the data
stream by

1.3 extracting a significance map indicating positions of significant

transform coefficients within the transform coefficient block, and
then the values of the significant transform coefficients within the
transform coefficient block from a data stream, with, in extracting
the significance map,

1.4 sequentially extracting first-type syntax elements from the data

stream by context-adaptive entropy decoding, the first-type
syntax elements indicating, for associated positions within the
transform coefficient block as to whether at the respective
position a significant or insignificant transform coefficient is
situated; and

1.5 sequentially associating the sequentially extracted first-type

syntax elements to the positions of the transform coefficient
block in a predetermined scan order among the positions of the
transform coefficient block,

1.6 using, in context-adaptively entropy decoding the first-type

syntax elements, contexts which are individually selected for
each of the first-type syntax elements depending on a number of
positions at which according to the previously extracted and
associated first-type syntax elements significant transform
coefficients are situated,

1.7 in a neighborhood of the position with which a current first-type

syntax element is associated.

5.2 Granted claim 3

71 Further, independent method claim 3 of the Patent is shown below,

divided into separate integers in the form of a feature analysis.

16
 3.1 Encoder for encoding an array of information samples into a data
stream in units of blocks, configured to encode blocks (40) of the
array of information samples by

3.2 using a transform in order to transfer the information samples

within each block (40) from spatial domain into spectral domain
in order to obtain transform coefficient blocks, and

3.3 inserting, by entropy encoding, information on the transform

coefficient blocks into the data stream by encoding a significance
map indicating positions of significant transform coefficients
within a transform coefficient block into the data stream, by

3.4 coding a significance map indicating positions of significant

transform coefficients within the transform coefficient block, and
then the values of the significant transform coefficients within the
transform coefficient block into the data stream, with, in coding
the significance map,

3.5 sequentially coding first-type syntax elements into the data

stream by context-adaptive entropy encoding, the first-type
syntax elements indicating, for associated positions within the
transform coefficient block as to whether at the respective
position a significant or insignificant transform coefficient is
situated,

3.6 the first-type syntax elements being sequentially coed into the
data stream in a predetermined scan order among the positions
of the transform coefficient block,

3.7 wherein, in context-adaptively entropy encoding each of the first-

type syntax elements, contexts are used which are individually
selected for the first-type syntax elements depending on a
number of positions at which significant transform coefficients
are situated and with which the previously coded first-type syntax
elements are associated,

3.8 in a neighborhood of the position with which a current first-type

syntax element is associated.

72 The feature analysis of both granted claim 1 and claim 3 is enclosed

herewith as

Annex M2.

17
C. Grounds for Opposition

I. Added subject matter – Articles 100 (c), 123 (2) EPC and Articles
100 (c), 76 EPC

73 Granted claim 1 differs from the originally filed claims in that granted claim
1 of the Patent does not contain the features regarding an associator as
disclosed in the originally filed and published claim 1 of the Patent
application.

74 Originally filed claim 1 of the Patent application further specifies that this
associator is sequentially associating the sequentially extracted first-type
syntax elements to the positions of the transform coefficient block in a
predetermined scan order among the positions of the transform coefficient
block.

75 This feature of the associator is in the granted claim of the Patent

attributed to the overall system, not to the associator as originally
disclosed, representing a so-called intermediate generalization, the
subject matter extending beyond the content of the application as
originally filed, contrary to the requirements of Article 123(2) EPC.

76 At least for these reasons, granted claim 1 extends beyond the content of
the application as originally filed, contrary to Article 123(2) EPC.

77 Therefore, the Patent is to be revoked in its entirety under Articles 100(c),

123(2) EPC.

78 Further, the subject matter regarding feature 1.5, e.g. “sequentially

associating the sequentially extracted first-type syntax elements to the
positions of the transform coefficient block in a predetermined scan
order among the positions of the transform coefficient block” extends
also beyond the content of the European Parent applications as
originally filed contrary to Article 76 EPC.

79 Thus, sequentially associating the sequentially extracted first-type syntax

elements as such is not disclosed in any one of the parent European
Parent applications.

80 Therefore, the Patent is to be revoked in its entirety under Articles 100(c),

76 EPC.

18
II. Lack of novelty – Articles 100 (a), 52 EPC

81 The subject matter of each of claims 1 to 9 of the Patent does not meet
the requirement of novelty over numerous prior art references, as
discussed in detail in the following.

1. Lack of novelty over cited document D2

82 In the following, detailed reasoning is provided why the features as

present in granted claim 1 of the Patent are disclosed by cited document
D2.

1.1 Independent claim 1

83 Independent claim 1 is not novel by D2.

84 To the extent of features 1.1 to 1.7, cited document D2 shows using a

significance map, entropy coding, and neighbourhood assigning.

85 Significance maps for decoding and encoding of data streams are also
known from for instance cited document D2, see for instance cited
document D2, page 609, section 5.15 discussing the CREW method,
corresponding to feature 1.3 of claim 1 of the Patent.

86 Cited document D2 also discusses entropy coding, thus feature 1.4 of

claim 1 of the Patent:

See page 630 of D2:

“Entropy Coding: The wavelet coefficients of a code-block are

arithmetically coded by bitplane. The coding is done from the most-
significant bitplane (containing the most important bits of the
coefficients) to the least-significant bitplane. Each bitplane is
scanned as shown in Figure 5.76. A context is determined for each
bit, a probability is estimated from the context, and the bit and its
probability are sent to the arithmetic coder.”

87 With regard to feature 1.5 and feature 1.6 of claim 1 of the Patent, it is
submitted that cited document D2 also discusses raster order scan on
pages 31, 259, 261, 275, 352, 395, 411, 416, 425–427, 437–439, 656,
663, 676.

88 With regard to feature 1.7 of the claim 1 of the Patent it is submitted that
cited document D2 discusses in Section 4.11 EIDAC, a method for

19
 compressing simple images. Its main goal is the use of two-part contexts.
The intra context of a pixel P consists of several of its near neighbors in
its bitplane. The inter context of P is made up of pixels that tend to be
correlated with P even though they are located in different bitplanes.

89 Thus feature 1.7 is disclosed by cited document D2.

90 In summary, the subject matter of claim 1 of the Patent as granted is

disclosed by cited document D2.

1.2 Independent claim 3

91 Independent claim 3 is disclosed by cited document D2, for multiple

reasons as stated above, in corresponding manner at least for the reasons
as provided above for granted claim 1 of the Patent since claim 3 of the
Patent is directed to the encoder corresponds to claim 1 directed to the
decoder as far as the features 3.1 to 3.8 of claim 3 of the Patent are
concerned, which clearly correspond to features 1.1 to 1.7 of claim 1 of
the Patent.

92 Thus, claim 1 and claim 3 of the Patent built a socket-plug

correspondence, apparently, with the lack of novelty over cited document
D2 for multiple reasons as stated above.

1.3 Independent claims 5 and 6

93 Independent claims 5 and 6 intrinsic difference between apparatus claims

1 and 3 on the one side and method claims 5 and 6 is given in an
apparatus and a method, respectively.

94 Thus, it remains to show that also a method is disclosed or at least

anticipated by the cited documents.

95 Cited document D2 teaches in Chapters 5, 7, and 8 a description of

important compression methods for encoding or decoding data streams.

96 Therefore, any one of method claims 5 and 6 lacks novelty over cited
document D2.

1.4 Independent claims 7 and 9

97 Claim 9 of the Patent recites as follows:

20
 “Computer readable digital storage medium having stored thereon
a computer program having a program code for performing, when
running on a computer, a method according to claims 5 or 6.

98 Cited document D2 teaches on page 860:

“Data Structure. A set of data items used by a program and stored

in memory such that certain operations (for example, finding,
adding, modifying, and deleting items) can be performed on the
data items fast and easily

99 Therefore, any one of claims 7 and 9 lacks novelty over cited document
D2.

III. Lack of inventive step – Articles 100 (a), 56 EPC

100 The subject matter of each of claims 1 to 9 of the Patent does not involve
inventive step over numerous prior art references, as discussed in detail
in the following.

1. Lack of inventive step over document D4 taken alone or in

combination with D5

1.1 Independent claim 1

101 Independent claim 1 is anticipated by D4 (“Lee”) taken also or in

combination with D5 (“Ye”), at least for the following reasons.

102 To the extent of features 1.1, 1.1, 1.2, 1.3, and 1.4, cited document D4,
“Lee” is analogous art to the Patent. The field of endeavor of the Patent
relates to adaptive coding of significance maps and transform coefficient
blocks for use in picture and video coding. See, e.g., the Patent at
paragraph [0001], Claim 1.

103 Like the Patent, D4 (“Lee”) also relates to and describes adaptive coding
and decoding of significance maps and transform coefficients for video
coding, see D4 (“Lee”) at [0003]; [0027]; see also Lippman Decl. D17 at
¶ 63.

104 Further, cited document D4 (“Lee”) is reasonably pertinent to at least one

problem with which the inventors of the Patent were concerned. For
example, the inventors of the Patent described that “the ability to
effectively code both small and large blocks would enable the

21
 achievement of better coding efficiency.” See the Patent at paragraph
[0011].

105 Thus, the inventors sought to address the problem of increasing coding
efficiency. cited document D4 (“Lee”) highlights the same problem of the
need to increase coding efficiency by stating that “a method of more
efficiently performing context modeling is needed.” See cited document
D4 (“Lee”) at [0026]; see also Lippman Declaration D17 at ¶ 62. Both Lee
and the patent are directed to solutions for increasing coding efficiency
using adaptive coding techniques. Lippman Declaration D17 at ¶¶61-63.

106 Cited document D5, (“Ye”), was filed on January 4, 2008 and claims
priority to provisional application No. 60/979762, filed on October 12,
2007. Cited document D5, (“Ye”), was published on December 18, 2008
as US Pat. Pub. 2008/0310745. Accordingly, cited document D5, (“Ye”),
qualifies as prior art as to the Patent.

107 Cited document D5, (“Ye”), is analogous art to the Patent and discusses
the same technical field of video coding. Like the Patent, cited document
D5, (“Ye”) also relates to and describes adaptive coefficient scanning for
the purposes of video compression.

108 See cited document D5, (“Ye”), at Summary, paragraph [0012]; see also
Lippman Decl. (D17) at ¶ 64.

109 Further, cited document D5, (“Ye”) is reasonably pertinent to at least one
problem with which the inventors of the Patent were concerned. For
example, Ye provides for techniques that utilize “thresholds and threshold
adjustments that can reduce the frequency that the scanning order
adjustments occur, yet still achieve desired improvements in compression
due to such scanning order adjustments.” cited document D5, (“Ye”), at
Summary, paragraph [0012].

110 As such, both cited document D5 (“Ye”) and the Patent are aimed at
increasing the coding efficiency of videos by utilizing adaptive scanning
techniques, allowing a reduction of the number of code bits used to
transmit the significance maps. Lippman Decl. (D17) at ¶¶ 61-62, 64.

111 As further explained below, a person skilled in the art would have been
motivated to combine cited document D4 (“Lee”) and cited document D5
(“Ye”) because cited document D5 (“Ye”) expressly considers the
traditional scan orders in cited document D4 (“Lee”), i.e. zig-zag scanning,
and improves on them.

22
112 Cited document D4 (“Lee”) taught the use of conventional zig-zag
scanning for scanning the coefficients within a sub-block. See e.g., cited
document D4 (“Lee”), at Fig. 7.

113 Cited document D5 (“Ye”) offered an improvement to this conventional

method through use of adaptive scan orders. “Rather than using
conventional zig-zag scanning, the scanning techniques described in this
disclosure adapt the scanning order.” See cited document D5 (“Ye”) at
[0012].

114 Similarly, it was well-known before the priority date of the Patent that
grouping zero value and non-zero value coefficients together in the scan
order increases coding efficiency as described in the background section
above.

115 The differing features of claim 1 of the Patent with regard to the cited
documents could be summarized as in that in the prior art the scan order
is not predetermined.

116 The problem allegedly solved by claim 1 can therefore be construed as to

reduce the number of syntax elements indicating as to whether at a
respective position a significant or insignificant transform coefficient is
situated.

117 For the person skilled in the art the solution to this problem would be
obvious starting for instance with cited document D5.

118 Regarding a predetermined scan order the person skilled in the art would
have considered the adaptive scan orders of D5.

119 Further, a person skilled in the art would have known the teachings of
cited document D5 (“Ye”) to incorporate this beneficial technique into the
system of cited document D4 (“Lee”) so that this well-known benefit could
be achieved in situations where significant coefficients were not
concentrated at the locations within blocks that are quickly reached and
grouped by conventional zig-zag scanning, as expressly described in
cited document D5 (“Ye”). Lippman Decl. D17 at ¶¶ 75-80.

120 Further, applying the adaptive scan order selection technique of cited
document D5 (“Ye”) to the coding method of cited document D4 (“Lee”)
would have been the combination of known prior art elements according
to known methods.

121 As a consequence, a person skilled in the art would have yielded the
predictable result of grouping together the zero and non-zero value
coefficients in a scan order to achieve the improved coding efficiency

23
 described in cited document D5 (“Ye”). This is because scan orders, and
adjustments to scan orders, were known to take advantage of
concentrations of zero and non-zero coefficients in a transform block,
which in turn improves coding efficiency.

122 Such a combination would have been routine to implement and would not
have required undue experimentation because techniques involving the
use of various scan orders, and specifically adaptive selection of scan
orders, are implemented in software and would not have required any
changes to the system architecture or hardware.

123 Thus, a person skilled in the art with even minimal experience in coding
could implement the required changes with reasonable efforts.

124 Furthermore, because such techniques were well-known at the time of the
Patent, the coding of such a change would not have required undue
experimentation and would have required little more than the
implementation of already existing routines, providing a high expectation
of success.

125 Therefore, claim 1 lacks inventive step over cited document D4 (“Lee”),
for multiple reasons as stated above, taken alone, or over cited document
D4 (“Lee”) in combination with cited document D5 (“Ye”).

126 As an alternative, claim 1 lacks inventive step over cited document D4

(“Lee”) in combination with common general knowledge, as common
general knowledge was discussed above in section titled “3.
Technological Background”.

127 Therefore, the Patent is to be revoked in its entirety under Articles 100(a),
54, 56 EPC.

128 Further, also cited document D3 could be considered for CABAC, entropy
coding, context modeling, and binary arithmetic coding, could be derived
from cited document D3, see page 1 and 2 of D3, see Fig. 1 of D3.

129 Thus, at least features 1.1, 1.2, 1.3, 1.4 could be derived from D3 by the
person skilled in the art. Context-based Arithmetic Encoding of 2D Shape
Sequences is also motivated by cited document D11, see abstract of D11.

130 In alternative, cited document D1 discusses significance mapping, relating

to features 1.3 and 1.4 of claim 1 of the Patent in section 1.3 Significance
mapping of D1.

131 With regard to feature 1.5 and the scan order, such scan orders are known
for the person skilled in the art for instance, zig-zag scanning as scan

24
 order, see in this regard for instance cited document D7 at page 10, “zig-
zag scanning pattern”, left column.

132 Cited document D8 provides hints for the person skilled in the art to
implement adaptive binary arithmetic coder as the in feature 1.6 of claim
1 of the Patent claimed context adaptively entropy decoding. See also
cited document 9, entropy of an information source, on page 45 of D9.

133 See also cited document D12 on page 16, D13 in particular the abstract
of D13, and D14, in particular introduction of D14, in this context.

1.2 Independent claim 3

134 Independent claim 3 is anticipated by and over cited document D4 (“Lee”),

for multiple reasons as stated above, taken alone, or in combination with
cited document D5 (“Ye”), in corresponding manner at least for the
reasons as provided above for granted claim 1 of the Patent.

135 This is because granted claim 3 of the Patent directed to the encoder
corresponds to granted claim 1 of the Patent directed to the decoder as
far as the features 3.1 to 3.8 are concerned. In other words, these features
3.1 to 3.8 of claim 3 of the Patent clearly correspond to features 1.1 to 1.7
of claim 1 of the Patent.

136 Thus, claim 1 and claim 3 of the Patent built a socket-plug

correspondence, apparently, with the lack of inventive step of the
decoding, the encoding also lacks inventive step over cited document D4
(“Lee”), for multiple reasons as stated above, taken alone, or in
combination with cited document D5 (“Ye”).

1.3 Independent claims 5 and 6

137 Independent claims 5 and 6 intrinsic difference between apparatus claims

1 and 3 on the one side and method claims 5 and 6 is given in an
apparatus and a method, respectively.

138 Thus, it remains to show that also a method is disclosed or at least

anticipated by the cited documents.

139 Cited document D5 (“Ye”) to the coding method of cited document D4

(“Lee”) both teach methods for encoding or decoding data streams.

25
 140 Therefore, any one of method claims 5 and 6 lacks inventive step over
cited document D4 (“Lee”), for multiple reasons as stated above, taken
alone, or in combination with cited document D5 (“Ye”).

1.4 Independent claims 7 and 9

141 Claim 9 of the Patent recites as follows:

“Computer readable digital storage medium having stored thereon

a computer program having a program code for performing, when
running on a computer, a method according to claims 5 or 6.

142 Cited document D4 (“Lee”) teaches a non-transitory computer-readable

storage medium). In particular, the description of paragraph [0087] of cited
document D4 (“Lee”) recites as follows, emphasis added:

“[0087] Accordingly, the techniques described herein may be

implemented in hardware, software, firmware, or any combination
thereof. Any features described as modules or components may
be implemented together in an integrated logic device or
separately as discrete but interoperable logic devices. If
implemented in software, the techniques may be realized at least
in part by a computer-readable medium comprising
instructions that, when executed, performs one or more of the
methods described above. The computer-readable data storage
medium may form part of a computer program product, which may
include packaging materials. The computer-readable medium may
comprise random access memory (RAM) such as synchronous
dynamic random access memory (SDRAM), read-only memory
(ROM), non-volatile random access memory (NVRAM),
electrically erasable programmable read-only memory
(EEPROM), FLASH memory, magnetic or optical data storage
media, and the like.”

143 Further, it is respectfully submitted that a computer-readable medium

comprising instructions for causing a programmable processor to perform
the methods as claimed by claim 5 or 6 of the Patent would also be
common general knowledge for the person skilled in the art.

144 It is obvious that if cited document D4 (“Lee”) discloses such program

code for performing, when running on a computer, the claimed method of
claim 5 or 6 of the Patent, as a consequence, also a data stream as claim
by claim 7 of the Patent is provided by the technical disclosure of cited
document D4 (“Lee”).

26
 145 Therefore, any one of claims 7 and 9 lacks inventive step over cited
document D4 (“Lee”), for multiple reasons as stated above, taken alone,
or in combination with cited document D5 (“Ye”).

1.5 Dependent claims

146 All the dependent claims 2, 4, 8 of the Patent, all relating to a depth map
content, are anticipated either by cited document D4 (“Lee”) taken alone
or by cited document D4 (“Lee”) in combination with cited document D5
(“Ye”), as discussed in the following.

1.5.1 Claim 2

147 All the dependent claims 2, 4, 8 of the Patent recite as follows:

“wherein the transform coefficient block relates to a depth map

content.”

148 As evident from paragraph [0006] of cited document D5 (“Ye”):

[0006]

Video compression generally includes spatial prediction and/or

temporal prediction. In particular, intra-coding relies on spatial
prediction to reduce or remove spatial redundancy between video
blocks within a given coded unit, which may comprise a video
frame, a slice of a video frame, or the like.

149 spatial information signals are processed. Claim 2 of the Patent defines
that the transform coefficient block relates to a depth map content.
However, the description of the Patent defines in paragraph [0011] of the
Patent that in general sample arrays are pictures or sample arrays
representing other spatially sampled information signals such as depth
maps or the like. Thus, a depth map is nothing else than spatially sampled
information signals.

150 Therefore, claim 2 lacks inventive step over cited document D5 (“Ye”).

IV. Sufficiency of Disclosure – Articles 100 (b), 83 EPC

151 As discussed in the following, the feature “sequentially extracting first-type

syntax elements from the data stream by context-adaptive entropy
decoding” of the Patent contain subject matter, which fails to meet the
requirements of Art. 83 EPC since the person skilled in the art is not

27
 enabled to carry out the invention based on the disclosure of the European
patent application documents of the Patent (see also T 487/89, T 297/90,
T 541/97).

152 In particular, the objected features of granted claim 1 of the Patent recite
as follows:

“sequentially extracting first-type syntax elements from the data

stream by context-adaptive entropy decoding.”

153 The Patent is directed to techniques for encoding and decoding digital
video data.

154 The only disclosure relating to “sequentially extracting” in the Patent

occurs in paragraph [0068], which recites as follows:"

“[0068] In other words, the decoder 250 may be configured to

sequentially extract the significance map syntax elements by
context-adaptively entropy decoding by use of contexts which are
individually selected for each of the significance map syntax
elements depending on a number of positions at which according
to the previously extracted and associated significance map syntax
elements significant transform coefficients are situated, the
positions being restricted to ones lying in a neighborhood of the
position ("x" in Fig. 9 right-hand side and Fig. 10 both sides, and
any of the marked positions of the left hand side of Fig. 9) with
which the respective current significance map syntax element is
associated.”

155 However, claim 1 is silent on the use of contexts which are individually
selected for each of the significance map syntax elements depending on
a number of positions at which according to the previously extracted and
associated significance map syntax elements significant transform
coefficients are situated.

156 In T 409/91 (OJ 1994, 653; ex parte) and T 435/91 (OJ 1995, 188; inter
partes) it was pointed out that the protection conferred by a patent should
correspond to the technical contribution to the art made by the disclosure
of the invention described therein, which excludes the patent monopoly
being extended to subject-matter which, after reading the patent
specification, would still not be at the disposal of the skilled person.

157 The general statement of sequentially extracting first-type syntax

elements from the data stream by context-adaptive entropy decoding is
clearly not at the disposal of the skilled person.

28
 158 The available information in paragraph [0069] of the Patent is too specific
and does not enable the skilled person to achieve the envisaged result
within the whole ambit of the claim containing the respective functional
definition without undue difficulty.

159 The technical teaching of paragraph [0069] of the Patent fails to provide
a fully self-sufficient technical concept as to how sequentially extracting
first-type syntax elements from the data stream by context-adaptive
entropy decoding as such is to be achieved.

160 Thus, in summary, the subject matter of each of the claims 1 and 3 and
corresponding methods claims 5 and 6 of the Patent fails to meet the
requirements of Article 100(b) EPC or Article 83 EPC, respectively.

D. Summary and Conclusion

161 It has been shown hereinabove that the claims as granted contain subject
matter extending beyond the content of application as originally filed as
well as the subject-matter of the European patent extends beyond the
content of the parent application as filed.

162 Therefore, the Patent is to be revoked in its entirety under Articles 100(c),
123(2) EPC and under Articles 100(c), 76 EPC.

163 Each one of the independent claims lacks novelty over the prior art.
Therefore, the Patent is to be revoked in its entirety under Articles 100(a)
EPC in conjunction with Articles 52(1) and 54(1),(2) EPC (lack of novelty)

164 In addition, each of the independent claims lacks inventive step over
numerous prior art documents.

165 Therefore, the Patent is to be revoked in its entirety under Articles 100(a),
54, 56 EPC.

29
 166 In addition, it has been shown hereinabove that the claims as granted
claim define an invention, which the person skilled in the art cannot carry
out based on the disclosure of the originally filed patent application
documents.

167 Therefore, the Patent is to be revoked in its entirety under Articles 100(b),
83 EPC.

Maiwald Patentanwalts- und Rechtsanwaltsgesellschaft mbH

- Association No. 174 -

electronically signed
Dr. Simon Q. Lud
Patentanwalt | European Patent Attorney

Enclosures
Annex M1 (list of references);
Annex M2 (feature analysis of claims 1 and 3);
Documents D1 to D17

30