Sie sind auf Seite 1von 15

Trials@uspto.

gov Paper 7
571-272-7822 Entered: December 9, 2019

UNITED STATES PATENT AND TRADEMARK OFFICE


____________

BEFORE THE PATENT TRIAL AND APPEAL BOARD


____________

UNIFIED PATENTS INC.,


Petitioner

v.

VELOS MEDIA, LLC,


Patent Owner
____________

Case IPR2019-01130
Patent 8,885,956 B2
____________

Before JENNIFER S. BISK, KIMBERLY McGRAW,


and AARON W. MOORE, Administrative Patent Judges.

MOORE, Administrative Patent Judge.

DECISION
Denying Inter Partes Review
35 U.S.C. § 314
Case IPR2019-01130
Patent 8,885,956 B2

I. INTRODUCTION

Unified Patents Inc. (“Petitioner”) filed a Petition for inter partes


review of claims 5 and 9–11 of U.S. Patent No. 8,885,956 B2 (Ex. 1001,
“the ’956 patent”). Paper 2 (“Pet.”). Velos Media, LLC (“Patent Owner”)
filed a Preliminary Response. Paper 6 (“Prelim. Resp.”).
Institution of an inter partes review is authorized by statute when “the
information presented in the petition . . . and any response . . . shows that
there is a reasonable likelihood that the petitioner would prevail with respect
to at least 1 of the claims challenged in the petition.” 35 U.S.C. § 314(a);
see 37 C.F.R. § 42.108.
Having considered the Petition, the Preliminary Response, and the
evidence of record, we conclude that Petitioner has not shown a reasonable
likelihood that it would prevail in establishing the unpatentability of any
challenged claim of the ’956 patent and, therefore, do not institute inter
partes review.

A. Related Matters

Petitioner identifies “United States Patent Application No.


15/707,509” as a related matter pending before the Office. See Pet. 49.
Patent Owner does not identify any related matters. See Patent Owner
Mandatory Notices (Paper 4) 2.

B. The ’956 Patent

The challenged claims are directed to video encoding and, more


specifically, to a technique called intra prediction. See Pet. 9. Intra
prediction is “designed to achieve compression by exploiting spatial
redundancies in a frame.” Id.

2
Case IPR2019-01130
Patent 8,885,956 B2

The ’956 patent describes the use of different intra prediction


“modes,” such as those for the 4x4 blocks illustrated in Figure 4:

“FIG. 4 is a diagram illustrating the kinds of 4x4 pixel intra


prediction modes for luminance signals.” Ex. 1001, 6:45–46.

The intra prediction modes use different neighboring pixels as


reference pixels. Mode 2 is a “DC Prediction Mode,” in which the
prediction is created from an average of the neighboring pixel values, while
the other modes are “Directional Prediction Modes” in which the prediction
is created by extrapolating neighboring pixel values in a certain direction,
indicated by the arrows. See Ex. 1002, 12:30–15:36.
The ’956 patent also explains that “with intra prediction according to
the H.264/AVC format, filtering processing of pixel values of neighboring
pixels is performed with determined filter coefficients only before

3
Case IPR2019-01130
Patent 8,885,956 B2

performing intra prediction in block increments of 8x8 pixels,” but that, in


the inventive method, “filtering processing of pixel values of neighboring
pixels is performed with filter coefficients set according to the block to be
predicted, prior to performing intra prediction of all intra prediction modes.”
Id. at 22:66–23:8.
Patent Owner acknowledges that the prior art H.264/AVC standard
used both intra prediction modes and filtering, but states that the inventors of
the ’956 patent recognized that the prior art approaches “did not take into
account numerous factors that should influence whether filter processing
should be performed, such as the degree of noise in the input picture and the
intra prediction mode.” Prelim. Resp. 8. The alleged improvement of the
’956 parent lies in a “neighboring pixel interpolation filter control unit” that
uses a “control signal” to “allow filter processing to be selectively turned
on/off for different intra prediction modes.” Prelim. Resp. 10.

C. The Challenged Claims

Claim 5 is the only independent claim being challenged and is, thus,
representative of the challenged claims:
5. An image processing method comprising the steps of:
performing, on the basis of an intra prediction mode, filter
processing turned on or off as to neighboring pixels that are
adjacently located to a current block for intra prediction,
which is to be the object of encoding processing, the
neighboring pixels subjected to the filter processing turned
on or off being used as reference pixels for the intra
prediction of the current block;
performing intra prediction on the current block using the
reference pixels to generate a prediction image; and
encoding an image of said current block using the prediction
image,

4
Case IPR2019-01130
Patent 8,885,956 B2

wherein the neighboring pixels are not included within the


current block,
wherein the filter processing changes value of respective pixels
being filtered,
wherein the filter processing on or filter processing off is
performed as to the neighboring pixels on the basis of a
mode number of the intra prediction mode, and
wherein the filter processing on or filter processing off is
performed on the basis of a control data indicating whether
or not to perform filter processing on the current block.
Ex. 1001, 76:60–77:16.

D. Asserted Grounds of Unpatentability

Petitioner asserts that claims 5 and 9–11 are unpatentable on the


following grounds:

References Basis Claims


Lee 1 and Karczewicz 2 § 103(a) 5 and 9–11
Lee and Dahlhoff3 § 103(a) 5 and 9–11

Petitioner also relies on a Declaration of Joseph P. Havlicek, Ph.D.,


filed as Exhibit 1002.

1
EP 1 845 732 A2 (Exhibit 1004).
2
US 2008/0260027 A1 (Exhibit 1005).
3
US 2004/0136458 A1 (Exhibit 1006).

5
Case IPR2019-01130
Patent 8,885,956 B2

II. DISCUSSION

A. Level of Skill in the Art

The level of skill in the art is a factual determination that provides a


primary guarantee of objectivity in an obviousness analysis. See Al-Site
Corp. v. VSI Int’l Inc., 174 F.3d 1308, 1323 (Fed. Cir. 1999) (citing Graham
v. John Deere Co., 383 U.S. 1, 17–18 (1966)). The level of skill in the art
also informs the claim construction analysis. See Teva Pharm. USA, Inc. v.
Sandoz, Inc., 135 S. Ct. 831, 841 (2015) (explaining that claim construction
seeks the meaning “a skilled artisan would ascribe” to the claim term “in the
context of the specific patent claim”).
Petitioner asserts that a person of ordinary skill in the art “would have
had at least a bachelor’s degree in electrical engineering or a closely related
scientific field, such as physics, computer science, or computer engineering,
or similar advanced post-graduate education in this area, with two years of
experience with video processing systems” and that “[a] person with less
education but more relevant practical experience, depending on the nature of
that experience and degree of exposure to image processing devices and
algorithms, could also qualify as a person of ordinary skill in the field of the
’956 patent.” Pet. 20–21 (citing Ex. 1002 (Havlicek) ¶ 29). Patent Owner
does not address this issue in the Preliminary Response.
Because Patent Owner does not dispute Petitioner’s characterization
of the level of ordinary skill in the art, and because we find it generally
consistent with the subject matter of the ’956 patent and cited references, we
adopt it for purposes of this analysis.

6
Case IPR2019-01130
Patent 8,885,956 B2

B. Claim Construction

In inter partes review proceedings based on petitions filed on or after


November 13, 2018, such as this one, we construe claims using the same
claim construction standard that would be used in a civil action under
35 U.S.C. § 282(b), as articulated in Phillips v. AWH Corp., 415 F.3d 1303
(Fed. Cir. 2005) (en banc), and its progeny. See 37 C.F.R. § 42.100(b).
Because Petitioner does not seek any claim constructions (see Pet. 21)
and Patent Owner does not mention the issue, and because we do not
otherwise see a need for formal claim construction to determine whether to
institute, we do not expressly construe any claim terms in this decision. See
Nidec Motor Corp. v. Zhongshan Broad Ocean Motor Co., 868 F.3d 1013,
1017 (Fed. Cir. 2017) (explaining that construction is needed only for terms
that are in dispute, and only as necessary to resolve the controversy).

C. The Prior Art

1. Lee
Lee describes “an apparatus for spatial prediction of image data” that
“includes a spatial prediction unit that spatially predicts pixel values of a
current block of an image using neighboring blocks in a row immediately
above the current block among neighboring blocks that are spatially adjacent
to the current block.” Ex. 1004 ¶ 15. The described “apparatus includes a
pixel value filtering unit 100, a prediction mode determination unit 120, and
a spatial prediction unit 140.” Id. ¶ 25.
The pixel value filtering unit “filters pixel values of neighboring
blocks in a row immediately above a current block . . . and outputs the
filtered pixel values to the prediction mode determination unit 120.” Ex.

7
Case IPR2019-01130
Patent 8,885,956 B2

1004 ¶ 26. The filtering is said to “prevent display quality degradation


caused by the spatial prediction using only the pixel values of the
neighboring blocks in the row immediately above the current block.” Id.
The prediction mode determination unit “determines a spatial
prediction mode for the current block using the pixel values of the
neighboring blocks in the row immediately above the 4x4 block and the
filtered pixel values obtained by the pixel value filtering unit 100 and
outputs the determined spatial prediction mode to the spatial prediction unit
140.” Ex. 1004 ¶ 29. Lee describes the available modes as follows:
The spatial prediction mode includes a first prediction mode
for prediction using the filtered pixel values of the neighboring
blocks in the row immediately above the 4x4 block in the
vertical direction, a second prediction mode for prediction
using the original pixel values of the neighboring blocks that
are not filtered in the vertical direction, a third prediction mode
for prediction using the filtered pixel values of the neighboring
blocks in the row immediately above the 4x4 block in the
diagonal down-right direction, and a fourth prediction mode
for prediction using the filtered pixel values of the neighboring
blocks in the row immediately above the 4x4 block in the
diagonal down-left direction. The spatial prediction mode may
further include a fifth prediction mode for prediction using the
original pixel values of the neighboring blocks that are not
filtered in the diagonal down-right direction and a sixth
prediction mode for prediction using the original pixel values
of the neighboring blocks that are not filtered in the diagonal
down-left direction.
Id. ¶ 33 (emphases added).
Essentially, Lee uses the neighboring blocks in the row immediately
above the 4x4 block, and each of Lee’s “modes” includes a direction and
whether the neighbor pixels are filtered or unfiltered.

8
Case IPR2019-01130
Patent 8,885,956 B2

2. Karczewicz
Karczewicz describes “techniques to efficiently signal the prediction
modes used to encode blocks in a macroblock.” Ex. 1005 ¶ 19. It explains
that “[i]n H.264-2005, a prediction mode is derived for each 4x4 block in a
16x16 macroblock” and “signaled from encoder to decoder for each block,
i.e., 16 modes per macroblock.” Id. ¶ 20.
Karczewicz’s improvement is that “a ‘mode uniformity’ indicator can
be signaled on the macroblock level to indicate whether all 4x4 blocks
within the macroblock are using the same prediction mode.” Id. ¶ 21. For
example, the mode uniformity indicator may be a bit, where “[i]f the bit is
true, the prediction mode is signaled only once for all the 4x4 blocks in a
given macroblock” and “[i]f the bit is false, the prediction mode is signaled
for each 4x4 block.” Id.

3. Dahlhoff
Dahlhoff describes a method in which “pixels on the edge of at least
one previously predicted neighboring picture block are . . . used to predict an
actual picture block” and “[i]n addition, those pixels that are used for
prediction . . . are pre-filtered along an edge of the current picture block to
be predicted.” Ex. 1006 ¶ 3. The reason for the pre-filtering is to address
the problem that when the conventional 4x4 directional prediction is scaled
to larger blocks, “very concise structures are produced in the predicted
picture block, . . . resulting in strong artifacts in the reconstructed picture.”
Id. ¶ 4.

9
Case IPR2019-01130
Patent 8,885,956 B2

D. Reasonable Likelihood of Success Analysis

Petitioner alleges that all challenged claims are unpatentable under


35 U.S.C. § 103(a) in view of Lee and Karczewicz and, separately, in view
of Lee and Dahlhoff. See Pet. 28–47. We conclude that Petitioner has not
made a showing sufficient to support institution.

1. The Lee and Karczewicz Ground

Petitioner argues that Lee discloses the subject matter of claim 5,


except that Lee “does not explicitly teach that ‘the filter processing on or off
is performed on the basis of a control data indicating whether or not to
perform filter processing on the current block.’” Pet. 36 (emphasis
omitted).4 Petitioner argues, however, that this feature would have been
obvious given the teachings of Karczewicz. See id. at 36–38, 39–44.
In particular, Petitioner points to Karczewicz’s mode uniformity
indicator, which it describes as “indicat[ing] to a decoder whether all blocks
in a macroblock are encoded according to the same intra prediction mode.”
Pet. 36. According to Petitioner:
Karczewicz’s “mode uniformity indicator,” is “control data
indicating whether or not to perform” intra prediction using a
specific mode for a “current block”—which can include any

4
One might argue Lee is sufficient to anticipate or, at least, render claim 5
obvious without more because the filtering choice of the modes could be
considered the claimed “control data,” or because it would have been
inherent or obvious to apply the mode filter choice using “control data.” We
are not free to consider such theories, however, because they were not
presented in the Petition. See Arthrex, Inc. v. Smith & Nephew, Inc., 935
F.3d 1319, 1326 (Fed. Cir. 2019) (explaining that the Board may not “craft
new grounds of unpatentability not advanced by the petitioner”); In re
NuVasive, Inc., 841 F.3d 966, 971–72 (Fed. Cir. 2016); In re Magnum Oil
Tools Int’l, Ltd., 829 F.3d 1364, 1381 (Fed. Cir. 2016).
10
Case IPR2019-01130
Patent 8,885,956 B2

4x4 block in a macroblock when the bit is set. . . . In other


words, when the bit (i.e., “control data”) is set to 1, all
“current block[s]”—i.e., blocks that will be subject to intra
prediction processing using neighboring reference pixels—
are encoded (and will thus be decoded) according to the
same intra prediction mode as the mode that is signaled for
the first block in the macroblock. . . . When combined with
the teachings of [Lee], the signaling of modes 1 or 2 along with
the mode uniformity indicator indicating that all blocks in a
macroblock are encoded with the same intra prediction mode,
indicates that the filtering is on or off, based on the mode
number signaled with the first block in the macroblock.
Pet. 37–38 (footnote and citations omitted).
Patent Owner argues that “the claimed ‘control data’ indicates
whether or not to perform filter processing prior to encoding an image of the
current block,” while “Petitioner’s allegations repeatedly acknowledge that
Karczewicz’s ‘mode uniformity indicator’ is a signal sent to a decoder.”
Prelim. Resp. 45. Patent Owner contends that “Karczewicz’s ‘mode
uniformity indicator’ cannot satisfy the claimed ‘control data’ because it
does not indicate a filtering (or other) decision prior to encoding” and that
the indicator is not “usable by an encoder prior to intra prediction, which is a
requirement of the claimed ‘control data.’” Id. at 46–47.
We agree with Patent Owner. Claim 5 requires that “the filter
processing on or filter processing off is performed on the basis of a control
data indicating whether or not to perform filter processing on the current
block.” The claimed “control data” thus must “indicat[e] whether or not to
perform filter processing on the current block.” As explained above,
however, Karczewicz’s “mode uniformity indicator” is simply a bit that can
be set in the first block of a macroblock to signal to a decoder that all of the
blocks in the macroblock use the same prediction mode. It does not indicate

11
Case IPR2019-01130
Patent 8,885,956 B2

“whether or not to perform filter processing,” or otherwise control the


method of encoding.
For this reason, and without reaching Patent Owner’s other
arguments, we conclude that Petitioner has not established that it would
prevail in showing that the challenged claims of the ’956 patent are
unpatentable as obvious in view of Lee and Karczewicz on the theory
articulated in the Petition.

2. Analysis of the Lee and Dahlhoff Ground


Petitioner alternatively relies on Dahlhoff for the claimed “control
data.” According to Petitioner, “Dahlhoff describes the benefits of filtering
blocks having a block edge length of greater than four pixels and explicitly
discloses that filtering is ‘on’ for certain block sizes and ‘off’ for other block
sizes.” Pet. 38 (citing Ex. 1006 ¶¶ 4, 21). Petitioner contends that
“Dahlhoff thus describes a system in which filtering ‘on’ and filter
processing ‘off’ is determined based on the size of the block edge” and
“[t]herefore, according to the teachings of Dahlhoff, if the edge length of the
current block is 4 or less, filtering should be ‘off,’ and if the edge length of
the current block is greater than 4, then filtering should be ‘on.’” Id. From
this, Petitioner concludes that “[a] POSA would have understood that an
indicator of block size, which signals whether filter processing is used or not
(i.e., ‘control data’)—would have been obvious in light of the disclosure of
Dahlhoff.” Id.
We do not agree with the analysis offered by Petitioner, for several
reasons.
First, we are not confident that Petitioner correctly interprets
Dahlhoff. As noted above, the reference describes filtering to avoid artifacts

12
Case IPR2019-01130
Patent 8,885,956 B2

caused by using 4x4 techniques on larger blocks. Dahlhoff explains that its
method only filters for large blocks when the edge of the large block is
greater than four pixels. See Ex. 1006 ¶ 21 (“Filtering is used at the edges
having lengths >4 pixels long, i.e., for edges that are 8 or 16 pixels long.”).
Dalhhoff is thus describing when to applying filtering for large blocks, not
that one should never filter 4x4 blocks, as Petitioner contends.
Second, even assuming the block size in Dahlhoff could fairly be
considered “control data,” it would be control data that determines whether
filter processing is on or off based on the size of the block, not “on the basis
of an intra prediction mode” as required by claim 5. Petitioner does not
adequately explain how use of the block size to control filtering would result
in a system like that of claim 5, in which the filter processing is performed
on the basis of an intra prediction mode.
Finally, we fail to find in Dahlhoff a teaching of “control data”
beyond what is already in Lee. Petitioner argues that the control data in
Dahlhoff is “an indicator of block size” or “the block size itself” (see
Petition 38–39; Ex. 1002 (Havlicek) ¶ 165), but the Petition does not show
how or where in Dahlhoff either of those are actually used to control
filtering. Dahlhoff thus provides nothing that is not already in Lee, in which
the modes include information about whether or not to perform filtering.
Given Petitioner’s position that Lee does not include “control data,” we are
unable to find that the analogous disclosure of Dalhhoff does.
For these reasons, and without reaching Patent Owner’s other
arguments, we conclude that Petitioner has not established that it would
prevail in showing that the challenged claims of the ’956 patent are
unpatentable as obvious in view of Lee and Dahlhoff.

13
Case IPR2019-01130
Patent 8,885,956 B2

III. CONCLUSION

Petitioner has not demonstrated a reasonable likelihood of prevailing


in showing the unpatentability of claims 5 and 9–11 of the ’956 patent.

IV. ORDER

In consideration of the foregoing, it is hereby:


ORDERED that an inter partes review is not instituted in this
proceeding.

14
Case IPR2019-01130
Patent 8,885,956 B2

FOR PETITIONER:

Andrew R. Sommer
Ashraf Fawzy
Roshan S. Mansinghani
sommera@gtlaw.com
afawzy@unifiedpatents.com
roshan@unifiedpatents.com

FOR PATENT OWNER:

Brent N. Bumgardner
Barry Bumgardner
Thomas C. Cecil
Matthew Juren
brent@nbafirm.com
barry@nelbum.com
tom@nbafirm.com
matthew@nelbum.com
DG-Velos-IPR@nbafirm.com

15

Das könnte Ihnen auch gefallen