Electronics and Telecommunications Research Institute

Patent Trials and Appeals Board

Oct 20, 2021

IPR2021-00827 (P.T.A.B. Oct. 20, 2021)

Trials@uspto.gov Paper 14
571-272-7822 Date: October 20, 2021

UNITED STATES PATENT AND TRADEMARK OFFICE

BEFORE THE PATENT TRIAL AND APPEAL BOARD

UNIFIED PATENTS, LLC,
Petitioner,

v.

ELECTRONICS AND TELECOMMUNICATIONS
RESEARCH INSTITUTE,
Patent Owner.

IPR2021-00827
Patent 9,781,448 B2

Before JAMESON LEE, SALLY C. MEDLEY, and
NATHAN A. ENGELS, Administrative Patent Judges.

LEE, Administrative Patent Judge.
DECISION
Denying Institution of Inter Partes Review
35 U.S.C. § 314

IPR2021-00827
Patent 9,781,448 B2
2
I. INTRODUCTION
A. Background and Summary
On April 22, 2021, Unified Patents, LLC (“Petitioner”) filed a petition
(Paper 1, “Petition”) requesting inter partes review of claims 1 and 3
(“challenged claims”) of U.S. Patent 9,781,448 B2 (Ex. 1001, “the ’448
patent”). Electronics and Telecommunications Research Institute (“Patent
Owner”) filed a preliminary response. Paper 7 (“Prelim. Resp.”). Petitioner
filed an authorized reply. Paper 11 (“Prelim. Reply”). Patent Owner filed
an authorized sur-reply. Paper 12 (“Prelim. Sur-Reply”).
An inter partes review may not be instituted unless “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). Further, a
decision to institute may not institute on fewer than all claims challenged in
the petition. SAS Inst., Inc. v. Iancu, 138 S. Ct. 1348, 1359–60 (2018). We
determine that Petitioner has not shown a reasonable likelihood of prevailing
on any claim.
Accordingly, the Petition is denied, and no inter partes review is
instituted.
B. Real Parties-in-Interest
Petitioner identifies only itself as real party-in-interest. Pet. 1, 1.
Patent Owner identifies itself, Dolby Laboratories, Inc., Dolby Laboratories
Licensing Corporation, and Access Advance LLC as real parties-in-interest.
Paper 4, 2.
Patent Owner contends that the Petition fails to name all real parties-
in-interest (“RPIs”) as required by 35 U.S.C. § 312(a)(2). Prelim. Resp. 72–
IPR2021-00827
Patent 9,781,448 B2
3
78. Patent Owner contends that Petitioner also should have named
Petitioner’s SEP Video Codec Zone members as RPIs. Id. Because we
determine Petitioner has not demonstrated a reasonable likelihood of
prevailing in establishing the unpatentability of any challenged claim, we do
not reach this issue.
C. Related Matters
Neither party identifies any related district court litigation or any
Board proceeding involving the ’448 patent. Paper 1, 1; Paper 4, 2.
D. The ’448 patent
The ’448 patent is directed to a method and apparatus for performing
intra-prediction by applying an adaptive filter to surrounding pixel values of
a block to be intra-predicted or the predicted pixel values of a current block.
Ex. 1001, 1:65–2:2. The ’448 patent states:
[T]here is provided a method of performing intra-prediction. The
method of performing intra-prediction includes determining
whether to apply a first filter to a reference pixel value based on
information about surrounding blocks of a current block, if, as a
result of the determination, the first filter is determined to be
applied, applying the first filter to the reference pixel values,
performing intra-prediction for the current block based on the
reference pixel value, determining whether to apply a second
filter to a prediction value for each prediction mode of the current
block, predicted by performing the intra-prediction based on the
information about the surrounding blocks, and if, as a result of
the determination, the second filter is determined to be applied,
applying the second filter to the prediction values for each
prediction mode.
Id. at 2:11–26. The method is not limited to application in an encoder or a
decoder. The ’448 patent further states:
In another aspect, there is provided an encoder. The
encoder includes a processor and memory connected to the
IPR2021-00827
Patent 9,781,448 B2
4
processor and configured to store pieces of information for
driving the processor. The processor is configured to determine
whether to apply a first filter to a reference pixel value based on
information about surrounding blocks of a current block, if, as a
result of the determination, the first filter is determined to be
applied, apply the first filter to the reference pixel value, perform
intra-prediction for the current block based on the reference pixel
value, determine whether to apply a second filter to a prediction
value for each prediction mode of the current block, predicted by
performing the intra-prediction based on the information about
the surrounding blocks, and if, as a result of the determination,
the second filter is determined to be applied, apply the second
filter to the prediction value for each prediction mode.
Id. at 2:54–3:3.
The ’448 patent still further states:
In yet another aspect, there is provided a decoder. The
decoder includes a processor and memory connected to the
processor and configured to store pieces of information for
driving the processor. The processor is configured to determine
whether to apply a first filter to a reference pixel value based on
information about surrounding blocks of a current block, if, as a
result of the determination, the first filter is determined to be
applied, apply the first filter to the reference pixel value, perform
intra-prediction for the current block based on the reference pixel
value, determine whether to apply a second filter to a prediction
value for each prediction mode of the current block, predicted by
performing the intra-prediction based on the information about
the surrounding blocks, and if, as a result of the determination,
the second filter is determined to be applied, apply the second
filter to the prediction value for each prediction mode.
Id. at 3:4–3:20.
Claims 1 and 3 are each independent and reproduced below:
1. A video decoding apparatus, comprising:
a processor to determine whether to apply a first filter to a
reference pixel value of a current block based on at least one
of an intra prediction mode of the current block and a size of
the current block,
IPR2021-00827
Patent 9,781,448 B2
5
apply the first filter to the reference pixel value if, as a result of
the determination, the first filter is determined to be applied,
and,
perform intra-prediction for the current block based on the
reference pixel value,
wherein when the first filter is determined to be applied,
the processor calculates a difference value using the
reference pixel values, compares the calculated
difference value with a specific threshold and applies
the first filter to the reference pixel value based on
comparison of the specific threshold and the calculated
difference,
when the intra-prediction for the current block is
performed, the processor determines whether to apply
a second filter to a prediction value of the current block
based on at least one of the intra prediction mode of the
current block and the size of the current block, and
applies the second filter to the prediction value of the
current block if, as a result of the determination, the
second filter is determined to be applied.
Ex. 1001, 12:2–25.
3. A video encoding apparatus, comprising:
a processor to determine whether to apply a first filter to a
reference pixel value of a current block based on at least one
of an intra prediction mode of the current block and a size of
the current block,
apply the first filter to the reference pixel value if, as a result of
the determination, the first filter is determined to be applied,
and,
perform intra-prediction for the current block based on the
reference pixel value,
wherein when the first filter is determined to be applied,
the processor calculates a difference value using the
reference pixel values, compares the calculated
difference value with a specific threshold and applies
the first filter to the reference pixel value based on
IPR2021-00827
Patent 9,781,448 B2
6
comparison of the specific threshold and the calculated
difference,
when the intra-prediction for the current block is
performed, the processor determines whether to
apply a second filter to a prediction value of the
current block based on at least one of the intra
prediction mode of the current block and the size of
the current block, and applies the second filter to the
prediction value of the current block if, as a result
of the determination, the second filter is determined
to be applied.
Id. at 12:30–53.
Claims 1 and 3 are identical, except that claim 1 in its preamble
recites “A video decoding apparatus,” and claim 3 in its preamble
recites “A video encoding apparatus.” Id. at 12:2, 12:30.
E. Evidence Relied on by Petitioner
Petitioner relies on the following evidence:1
References Date Exhibit
Moon U.S. Pub. App. US
2005/0201633 A1
Published Sep. 15, 2005 Ex. 1004
Teng U.S. Pat. 8,681,867 B2 Issued Mar. 25, 2014,
filed Feb. 8, 2006
Ex. 1005
Wilkins U.S. Pat. 8,326,075 B2 Issued Dec. 4, 2012, filed
Dec. 5, 2008
Ex. 1006

1 The ’448 patent issued from Application 14/825,825, filed Aug. 13, 2015,
which is a continuation of Application 13/640,014, filed as
PCT/KR2011/002514, on April 11, 2011, now Patent 9,549,204. Ex. 1001
at codes (21), (22), (63). The ’488 patent also claims foreign priority to KR
10-2010-0032778, filed Apr. 9, 2010, KR 10-2011-0026079, filed Mar. 23,
2011, and KR 10-2011-0032766, filed Apr. 8, 2011. Id. at code (30).

IPR2021-00827
Patent 9,781,448 B2
7
Petitioner also relies on the Declaration of Immanuel Freedman, Ph.D.
(Ex. 1003).
F. Asserted Grounds of Unpatentability
Petitioner asserts that the challenged claims are unpatentable on the
following grounds:
Claims Challenged 35 U.S.C. §2 Basis
1, 3 103 Moon, Wilkins, Teng
1, 3 103 Moon, Teng
II. ANALYSIS
A. Level of Ordinary Skill
Petitioner asserts that one of ordinary skill in the art as of April 9,
2010, would have had “(1) at least an undergraduate degree in electrical
engineering or computer engineering or a closely related scientific field,
such as physics or computer science, or similar advanced post-graduate
education in this area; (2) a working knowledge of video coding techniques;
and (3) at least two years of experience with video processing.” Pet. 10
(citing Ex. 1003 ¶¶ 31–32). Patent Owner has not made its own assertion of
the level of ordinary skill. Nor has Patent Owner expressed disagreement
with Petitioner’s articulation.

2 The Leahy-Smith America Invents Act (“AIA”), Pub. L. No. 112–29, 125
Stat. 284, 287–88 (2011), revised 35 U.S.C. §§ 102 and 103 effective
March 16, 2013. Because the effective filing date of the ’488 patent is prior
to March 16, 2013 (Ex. 1001, codes (22), (63)), we refer to the pre-AIA
versions of §§ 102, 103.
IPR2021-00827
Patent 9,781,448 B2
8
Petitioner’s use of the qualifier “at least” renders the articulation
vague and potentially overlapping with the level of skill of an expert. Also,
“similar advanced post-graduate education” is not consistent with an
undergraduate degree first introduced in Petitioner’s articulation.
On this record, we adopt Petitioner’s articulation of the level of
ordinary skill, but delete the qualifier “at least” for the level of education and
for the amount of experience, to keep the level from being vague and
overbroad and extending into the level of an expert, and delete “or similar
advanced post-graduate education in this area” also to keep the level from
potentially extending into the realm of an expert. What remains appears
consistent with what is reflected by the content of the applied prior art
references. Cf. Okajima v. Bourdeau, 261 F.3d 1350, 1354–55 (Fed. Cir.
2001) (the applied prior art may reflect an appropriate level of skill).
For purposes of this Decision, we find the level of ordinary skill in the
art corresponds to that of one who had, on or about April 9, 2010, the level
of: (1) an undergraduate degree in electrical engineering or computer
engineering or a closely related scientific field, such as physics or computer
science; (2) a working knowledge of video coding techniques; and (3) two
years of experience with video processing.
B. Claim Construction
We use the same claim construction standard that would be used to
construe the claim in a civil action under 35 U.S.C. § 282(b), including
construing the claim in accordance with the ordinary and customary
meaning of such claim as understood by one of ordinary skill in the art and
the prosecution history pertaining to the patent. 37 C.F.R. § 42.100(b)
(2020). The Petition here was filed on April 22, 2021. Paper 1. We apply
IPR2021-00827
Patent 9,781,448 B2
9
the claim construction standard from Phillips v. AWH Corp., 415 F.3d 1303
(Fed. Cir. 2005) (en banc) (“the Phillips standard”).
Claim terms are generally given their ordinary and customary
meaning as would be understood by one with ordinary skill in the art in the
context of the specification, the prosecution history, other claims, and even
extrinsic evidence including expert and inventor testimony, dictionaries, and
learned treatises, although extrinsic evidence is less significant than the
intrinsic record. Phillips, 415 F.3d at 1312–17. Usually, the specification is
dispositive, and it is the single best guide to the meaning of a disputed term.
Id. at 1315.
The specification may reveal a special definition given to a claim term
by the patentee, or the specification or prosecution history may reveal an
intentional disclaimer or disavowal of claim scope by the inventor. Id. at
1316. If an inventor acts as his or her own lexicographer, the definition
must be set forth in the specification with reasonable clarity, deliberateness,
and precision. Renishaw PLC v. Marposs Societa’ per Azioni, 158 F.3d
1243, 1249 (Fed. Cir. 1998). The disavowal, if any, can be effectuated by
language in the specification or the prosecution history. Poly-America, L.P.
v. API Indus., Inc., 839 F.3d 1131, 1136 (Fed. Cir. 2016).
Only those claim terms that are in controversy need to be construed,
and only to the extent necessary to resolve the controversy. Nidec Motor
Corp. v. Zhongshan Broad Ocean Motor Co., 868 F.3d 1013, 1017 (Fed.
Cir. 2017); Wellman, Inc. v. Eastman Chem. Co., 642 F.3d 1355, 1361 (Fed.
Cir. 2011); Vivid Techs., Inc. v. Am. Sci. & Eng’g, Inc., 200 F.3d 795, 803
(Fed. Cir. 1999).
Petitioner has not proposed a construction for any claim term.
However, in applying the prior art to the claims, Petitioner reads the claim
IPR2021-00827
Patent 9,781,448 B2
10
term “prediction value,” appearing in both claims 1 and 3, onto original
values which have been reconstructed or restored based in part on values
which have been earlier predicted, and original values, of course, are not
themselves predicted. Patent Owner proposes an express construction for
“prediction value.” Prelim. Resp. 8–16.
A petition is required to set forth “[h]ow the challenged claim is to be
construed.” 37 C.F.R. § 42.104(b)(3). Although that does not mean
Petitioner must expressly construe each claim term, in this case Petitioner
should have provided and explained its own construction of “prediction
value,” a critical term in the claims.
“prediction value”
Patent Owner asserts that “prediction value” should be construed to
mean “a value generated in an intra-prediction process for determination of a
prediction block.” Prelim. Resp. 8. We agree.
Claims 1 and 3 each recite:
when the intra-prediction for the current block is
performed, the processor determines whether to apply a second
filter to a prediction value of the current block . . . , and applies
the second filter to the prediction value of the current block if,
as a result of the determination, the second filter is determined
to be applied
Ex. 1001, 12:18–25, 12:46–53 (emphasis added). The language links
“prediction value” to intra-prediction such that it is a predicted value
generated inside an intra-prediction process performed on a current block to
determine a prediction block for the current block.
The Specification of the ’448 patent supports this reading by
conveying that intra-prediction concludes with the final determination and
outputting of prediction block “P” from the box in Figure 1 that is labeled
IPR2021-00827
Patent 9,781,448 B2
11
“Intra-prediction.” Ex. 1001, 5:6–6:4, Fig. 1. For clarity, we reproduce
Figure 1 below, with color annotation (“Intra-prediction” box circled in red,
and prediction block “P” colored in red with textual annotation):

The above Figure is a color annotated version of Figure 1 of the ’448 patent,
which is a block diagram of a described encoder. Ex. 1001, 3:30–31.
We agree with Patent Owner that “prediction value” means “a value
generated in an intra-prediction process for determination of a prediction
block.” Values generated in post intra-prediction stages which use the
prediction block as partial input, such as (1) residual or difference block Dn
calculated by subtracting prediction block P from the current block (Ex.
1001, 4:33–34), or (2) reconstruction block uF′n calculated by adding
prediction block P back to restored difference macroblock D′n (Ex. 1001,
4:43–54), are not and cannot be prediction values. Residual or difference
IPR2021-00827
Patent 9,781,448 B2
12
block Dn and reconstruction block uF′n are generated outside of the intra-
prediction process while using completely determined prediction block P as
partial input. Indeed, reconstruction block uF′n is a restored approximation
of the original block Fn as it existed prior to the performance of any intra
prediction. Ex. 1001, 4:52–55; see also Ex. 1009, 186.3
The Specification further supports this construction by not referring to
anything subsequent to the production of prediction block P as a part of intra
prediction. We agree with Patent Owner that the ’448 patent makes clear
that intra prediction for a current block does not extend beyond computation
of the final prediction block P. Prelim. Resp. 12–13. The ’448 patent states:
“If, as a result of the determination, the filter is determined to have been
applied to the prediction value, the encoder applies the filter to the prediction
value at step S205. Accordingly, the prediction of the current block is
completed, . . . .” Ex. 1001, 9:13–17.
For this decision, no other term requires an express construction.4
C. Alleged Obviousness of
Claims 1 and 3 over Moon, Wilkins, and Teng
1. Law on Obviousness
A claim is unpatentable under 35 U.S.C. § 103 if the differences
between the claimed subject matter and the prior art are such that the subject
matter, as a whole, would have been obvious to a person having ordinary
skill in the art to which said subject matter pertains. See KSR Int’l Co. v.
Teleflex Inc., 550 U.S. 398, 406 (2007). The question of obviousness is

3 Iain E.G. Richardson, H.264 and MPEG-4 Video Compression, John Wiley
& Sons Ltd. (2003).
4 We need not determine whether the preamble of claims 1 and 3 is limiting
because, as discussed below, Petitioner has not adequately accounted for the
limitations in the body of claims 1 and 3.
IPR2021-00827
Patent 9,781,448 B2
13
resolved on the basis of underlying factual determinations, including (1) the
scope and content of the prior art; (2) any differences between the claimed
subject matter and the prior art; (3) the level of skill in the art; and (4) where
in evidence, so-called secondary considerations, including commercial
success, long-felt but unsolved needs, failure of others, and unexpected
results (“the Graham factors”). Graham v. John Deere Co. of Kansas City,
383 U.S. 1, 17–18 (1966).
2. Overview of Moon
Moon relates to encoding and decoding of motion picture data and a
filter for removing a blocking effect. Ex. 1004 ¶ 3. Moon explains that
because coding is performed in block units, a blocking effect arises where
discontinuity occurs at boundaries between blocks. Id. ¶ 6. Moon states that
a blocking effect is a type of image quality degradation, where discontinuity
on block boundaries occurs regularly like laid tiles as compression rate
increases. Id. ¶ 9. Moon describes that to remove such blocking effect,
filters are used and filters are classified into loop filters and post filters. Id.
Loop filters output a filtered frame which is then used as a reference frame
for the next frame. Id. ¶ 10. According to Moon, “[p]ost filters are located
on the rear portions of encoders and can be designed independently of
decoders.” Id.
Figure 1 of Moon shows a block diagram of an encoder and is
reproduced below:
IPR2021-00827
Patent 9,781,448 B2
14

Figure 1 illustrates an encoder including motion estimator unit 102, motion
compensator 104, intra predictor 106, transformer 108, quantizer 110, re-
arranger 112, entropy decoder 114, de-quantizer 116, inverse transformer
118, filter 120, and frame memory 122. Id. ¶ 39. The input to filter 20 is a
reconstructed or restored block uF′n – not any output of intra predictor 106.
Id. ¶ 48. The filtered restored picture is then stored in frame memory 122
“and is then used to perform an interprediction on a picture that follows the
current picture.” Id.
Moon also describes, regarding filter 120, selectively applying the
filtering depending on the circumstance. Id. ¶¶ 56, 85–87. Moon describes
that whether to apply filtering depends on a boundary strength (Bs
parameter) of the block. Id. ¶¶ 85–87. Moon provides a table showing
conditions that correspond to various Bs parameters:
IPR2021-00827
Patent 9,781,448 B2
15

Id. ¶ 85. Moon states that a “Bs parameter of 0 indicates that there is no
need for filtering.” Id. ¶ 87.
3. Overview of Teng
Teng explains: “[W]hen a video frame is divided into video blocks
for video coding, the edge of one coded video block may appear
discontinuous with the adjacent edge of another coded video block. When
this occurs, the decoded video frame may appear ‘blocky,’ which is highly
undesirable.” Ex. 1005, 2:5–8. Teng further explains:
In order to remove such “blockiness,” filtering can be
performed on the block boundaries of the video blocks to
“smooth” the transitions between adjacent video blocks.
Deblocking filters generally refer to filters that are used to
smooth the transitions between adjacent video blocks to reduce
or eliminate blockiness artifacts.
Ex. 1005, 2:11–16. Teng states that it provides for “selective deblock
filtering.” Id. at 2:42. Teng further describes that deblock filtering may be
selectively performed while a video block is being coded, after a video
frame is predicted, or both. Id. at 2:16–21, 2:48–52. Teng, however, does
not describe that post filtering is applied to obtain final output values of intra
prediction. Figure 1 illustrates an exemplary video coding device of Teng
and is reproduced below:
IPR2021-00827
Patent 9,781,448 B2
16

Figure 1 is a block diagram of Teng’s video coding device 10. Id. at
5:35. Teng states: “Coding device 10 may comprise any of a wide variety
of devices that can be used to encode or decode video sequences.” Id. at
5:36–38. Teng further describes: “Examples of coding device 10 generally
include any computer, such as a server, a workstation or any other desktop
computing device, or a mobile computing device such as a laptop computer
or a personal digital assistant (PDA).” Id. at 5:38–42.
Teng describes that its techniques may be implemented in hardware,
software, firmware, or any combination thereof. Id. at 14:10–11. Teng
further describes: “The program code may be stored on memory in the form
of computer readable instructions. In that case, a processor such as a DSP
may execute instructions stored in memory in order to carry out one or more
of the deblock filtering techniques.” Id. at 14:22–25.
4. Overview of Wilkins
Wilkins discloses a loop filter that removes or reduces blocking
artifacts. Ex. 1006, 2:41–42. Loop filter 34 is shown in Figures 1 and 2. Id.
IPR2021-00827
Patent 9,781,448 B2
17
at 2:41–45, 4:13–14, Fig. 1, 2. The output of loop filter 34 is provided to
intra/inter prediction stage 18 (Figure 1) and intra/inter prediction stage 23
(Figure 2) for performing intra/inter prediction. Id. at Figs. 1, 2.
Wilkins also discloses a post-loop filter. Id. at 4:14–17; Fig. 2
(deblocking filter 33). Deblocking filter 33, however, is disposed
subsequent to reconstruction stage 31. Id. at 4:10–17; Fig. 2. Wilkins states:
“At the reconstruction stage 31, the prediction macroblock can be added to
the derivative residual to create a reconstructed macroblock.” Id. at 4:10–
13. Thus, it is not the prediction macroblock from intra/inter prediction
stage 23 that is provided to deblocking filter 33, but a reconstructed
macroblock from reconstruction stage 31. Id. at 4:10–17, Fig. 2. Figure 2 of
Wilkins shows a video decompression system (id. at 2:22–23) and is
reproduced below:

Figure 2 shows decoder 21 which has a structure similar to the
reconstruction path of encoder 14 of Figure 1 (not shown). Id. at 3:59–67.
Further, Wilkins describes modifying the strength of a loop filter, e.g.,
loop filter 34. Id. at 3:1–3. In that regard, Wilkins states: “A filter that is
too strong may cause blurring and loss of detail. A filter that is too weak
IPR2021-00827
Patent 9,781,448 B2
18
may not adequately suppress discontinuities between adjacent blocks.” Id.
at 3:3–6. Specifically, Wilkins discloses application of strength modifier 60
to provide additional loop filter control, based in part on whether intra
prediction of the current block is performed in “split mode” or non “split
mode.”5 Id. at 2:27–31, Fig. 4, 5.
5. Petitioner’s Combination of Moon, Teng, and Wilkins
Petitioner asserts that the disclosures of Moon, Teng, and Wilkins all
relate to both encoding and decoding. Id. at 21 (citing Ex. 1004 ¶ 3; Ex. 1005,
6:10–13; Ex. 1006, 3:7–10, 3:59–63). Petitioner explains:
The proposed combination relies on the video coding
apparatus of Moon, modified by the addition of Wilkins’s
additional filtering criterion and in-loop deblocking filter. In
particular, in the proposed combination, Moon’s filtering criteria
(as embodied in filtering decision Table 1) is augmented by the
additional filtering criteria of Wilkins based on the size of the
current block and used to selectively apply both the post-loop
filter of Moon and the in-loop filter of Wilkins. The video coding
apparatus would include a processor such as Teng’s processor to
perform the filter decision. Teng’s processor is particularly
appropriate for implementing the conditional filtering decision
because Teng’s disclosed processor is programmed to determine
whether to apply a deblocking filter (such as, in the proposed
combination, based on the criteria of Moon and Wilkins).
Pet. 20. Petitioner further explains: “As applied to the system of Moon,
such a modification would result in an additional (conditional) filter being
applied to the reconstructed block before it is used for the purpose of intra-
prediction, as shown in the diagram below from Dr. Freedman’s Declaration.
Freedman Decl. at ¶ 55.” Id. at 22. The referenced diagram illustrating
what results from Petitioner’s proposed combination is reproduced below:

5 Dr. Freedman testifies that a distinction based on split-mode or non-split
mode equates to a distinction based on block size. Ex. 1003 ¶ 53.
IPR2021-00827
Patent 9,781,448 B2
19

The Figure is Petitioner’s depiction of the resulting structure based on its
proposed combination of Moon, Teng, and Wilkins. Pet. 22–23.
6. Apply a filter to a prediction value of the current block
Both claims 1 and 3 recite a processor that “when the intra-prediction
for the current block is performed, . . . determines whether to apply a second
filter to a prediction value of the current block” and “applies the second filter
to the prediction value of the current block if, as a result of the
determination, the second filter is determined to be applied.” Ex. 1001,
2:18–25, 2:46–53. In Section II.B., we have construed “prediction value” to
mean “a value generated in an intra-prediction process for
determination of a prediction block.”
Petitioner identifies Moon’s filter 120, i.e., filter 120 in the proposed
combination of Moon, Teng, and Wilkins in the immediate above-
reproduced Figure, as the claimed second filter. Pet. 47–48. Moon’s filter
120, however, does not filter any prediction value of the current block.
Rather, what is filtered by Moon’s filter 120 is reconstructed or restored
block uF′n. Ex. 1004, Fig. 1. In Petitioner’s proposed combination of Moon,
Teng, and Wilkins, what is filtered by filter 120 is version of reconstructed
IPR2021-00827
Patent 9,781,448 B2
20
or restored block uF′n which has been filtered once through a loop filter
added on the basis of Wilkins (shown in the immediate above-reproduced
Figure in red dotted lines). Pet. 22–23. As explained above in our claim
construction analysis in Section II.B., however, a reconstructed or restored
block (filtered or not filtered) does not include prediction values for the
current block. In Moon, uF′n is generated at a stage in the coding process
subsequent to the generation and complete determination of prediction block
P. Ex. 1004, Fig. 1. Indeed, Moon states: “The restored current picture is
stored in the frame memory 122, and is then used to perform an inter-
prediction on a picture that follows the current picture. If the restored
picture passes through the filter 120, it becomes the original picture that
additionally includes several encoding errors.” Id. ¶ 48 (emphasis added).
Petitioner makes no explanation as to why reconstructed or restored
block uF′n would include prediction values for the current block, such that
filter 120 is applied to a “prediction value” of the current block. Petitioner
has not made a sufficient showing for a processor that “when the intra-
prediction for the current block is performed, . . . determines whether to
apply a second filter to a prediction value of the current block” and “applies
the second filter to the prediction value of the current block if, as a result of
the determination, the second filter is determined to be applied.”
7. Based on at least one of an intra prediction mode of the
current block and size of the current block
Both claims 1 and 3 recite “a processor to determine whether to apply
a first filter to a reference pixel value of a current block based on at least
one of an intra prediction mode of the current block and a size of the current
block.” Ex. 1001, 12:3–6 (emphasis added). Both claims 1 and 3 also recite
“the processor determines whether to apply a second filter to a prediction
IPR2021-00827
Patent 9,781,448 B2
21
value of the current block based on at least one of the intra prediction mode
of the current block and the size of the current block.” Id. at 12:19–22
(emphasis added).
For reasons discussed below, Petitioner has not made a sufficient
showing for either condition on which application of the first and second
filter is based, i.e., (1) an intra prediction mode of the current block, and
(2) a size of the current block.
a) based on an intra prediction mode of the current block
The ’448 patent explains that there are nine different intra prediction
modes for intra predicting a 4x4 block, and four intra prediction modes for
intra predicting a 16x16 macroblock. Id. at 5:7–19. For example, the nine
different modes for intra predicting a 4x4 block are illustrated in Figure 4,
reproduced below:

IPR2021-00827
Patent 9,781,448 B2
22
Figure 4 shows nine intra prediction modes for a 4x4 block. Id. at 3:36–37,
5:61–6:4.
For the condition that application of the first and second filter is based
on an intra prediction mode of the current block, Petitioner relies on the
disclosure of Moon. Pet. 16–17. Specifically, Petitioner asserts:
Moon further teaches that deblocking filters are selectively
applied based on a set of criteria that involves calculating a
boundary strength parameter, Bs, which is calculated, in part,
based on the intra prediction mode, as shown in Table 1 below:

Moon at [0085]-[0087]. As shown in Table 1, the Bs parameter
is calculated based on the intra prediction mode, and whether
blocks are located on the boundary of a macroblock, along with
the residual signal and the motion compensation parameters.
According to Moon, if the Bs value is greater than 0 (i.e., 1, 2,
3, or 4), filtering is applied with varying strength. Id. at [0087].
If the Bs value is equal to 0, no filtering is applied. Id. (“The
Bs parameter of 0 indicates that there is no need for filtering.”).
Thus, Moon teaching selectively applying deblocking filters
based on a set of criteria that includes the intra prediction mode.
Id.
We are unpersuaded that Moon sufficiently teaches determining
whether to apply a deblocking filter based on an intra prediction mode of the
current block. The table referred to by Petitioner and reproduced above
makes no distinction among multiple intra prediction modes that may be
applied to a current block. In other words, based on this table, all intra
IPR2021-00827
Patent 9,781,448 B2
23
prediction modes are equivalent and they do not lead to a difference in
whether to apply a deblocking filter.
b) based on a size of the current block
To satisfy determining whether to apply a filter based on a size of the
current block, Petitioner relies on Wilkins’s disclosure of determining a filter
strength modifier 60 based in part on whether a block is processed in “split
mode” or “non-split mode.” Pet. 36–38. Wilkins explains that a macroblock
can be processed as a 16x16 block (non-split mode) or 16 4x4 sub-blocks
(split mode). Ex. 1006, 5:41–43. On that basis, Petitioner asserts that
“Wilkins’s split mode [and non-split mode] thus indicates a block size.” Pet.
36. Also on that basis, Petitioner asserts that “Wilkins teaches the filter
modifiers provide additional control of the filter strength as a function of
block size.” Id. at 28. However, for reasons discussed below, we are not
sufficiently persuaded by Petitioner’s assertion that in Wilkins filtering is
selectively applied based on a size of the current block.
Wilkins consistently indicates that filtering decisions are applied at
the macroblock level. Ex. 1006, 6:39–51, 2:49–51, Fig. 5. For instance,
Wilkins states:
As illustrated in FIG. 5, for example, a different strength
modifier 60 can be applied to intra-coded macroblocks rather
than inter-coded macroblocks (i.e. F1-F2 vs. F3-F11). Further,
a different strength modifier 60 can be applied to intra-coded
non-split mode macroblocks rather than intra-coded split-mode
macroblocks (i.e. F1 vs F2).
Id. at 8:2–9 (emphasis added). Wilkins further describes: “In the disclosed
embodiments, block-based video compression operates on fixed-shape
groups of neighboring pixels, called a macroblock.” Id. at 2:49–51. Wilkins
additionally states: “To adjust [filter] strength modifier 60 at the
IPR2021-00827
Patent 9,781,448 B2
24
macroblock level, delta values 1–8 can be encoded in the bitstream.” Id.
at 6:39–40. On this record, we agree with Patent Owner that:
Wilkins always applies a corresponding filter strength modifier
to the macroblock as a whole. Wilkins never tailors filtering
criteria such as strength modifiers to individual sub-blocks
within macroblocks, and never adjusts a strength modifier based
on the “size” of the block that is to be filtered (since Wilkins only
determines strength modifiers for entire macroblocks, which
have a fixed size).
Prelim. Resp. 54. Similarly, on this record, we agree with Patent Owner that
“the size of the ‘current block’ to which Wilkins applies the filtering criteria
in Wilkins’s Figure 5 is always the same, namely the fixed size of the
‘macroblock.’” Id. at 55. Petitioner has not sufficiently shown that in
Wilkins a different choice of filtering or not filtering is adopted for a 4x4
sub-block than for a 16x16 megablock.
8. Wilkins’s split mode / non-split mode
identification does not determine whether to apply a filter
As noted above, to satisfy determining whether to apply a filter based
on a size of the current block, Petitioner relies on Wilkins’s disclosure of
determining a filter strength modifier 60 based in part on whether a block is
processed in “split mode” or “non-split mode.” Pet. 36–38. Specifically, the
Petition states: “[I]n the proposed combination, Moon’s filtering criteria (as
embodied in filtering decision Table 1) is augmented by the additional
filtering criteria of Wilkins based on the size of the current block and used to
selectively apply both the post-loop filter of Moon and the in-loop filter of
Wilkins.” Id. at 20 (emphasis added).
We are not sufficiently persuaded by Petitioner’s assertion that a
person of ordinary skill in the art would have augmented Moon’s table of
filtering criteria to include Wilkins’s identification of the “split mode” or
IPR2021-00827
Patent 9,781,448 B2
25
“non-split mode” condition. As noted above, Moon states, with respect to its
Table 1, that a “Bs parameter of 0 indicates that there is no need for
filtering.” Ex. 1004 ¶ 87. Where “Bs” is greater than zero, depending on
other criteria, filtering still may not be applied. Id. ¶¶ 90–92. Wilkins’s
filtering scheme as shown in its Figure 5, however, is different in kind. At
the outset, Wilkins’s scheme begins with the recognition and acceptance that
a baseline loop filter of certain strength “f” is applied for filtering. Ex. 1006,
Figure 5 (box 100). Wilkins states:
Referring to Fig. 5, at block 100, a baseline loop filter strength f
can be selected for the frame that defines the behavior of adaptive
loop filter 34. Accordingly, baseline filter strength f will be
specified at the frame level in the encoded bitstream. By
specifying the baseline filter strength f at the frame level,
overhead can be reduced since very few bits can be used to
specify a single baseline filter value f for a whole frame.
Id. at 6:27–34. Wilkins further explains that to provide a filter strength
modifier at the macroblock level, delta values 1–8 can be encoded in the
bitstream, and then added to baseline filter strength f. Id. at 6:39–41.
Wilkins further describes: “Application of the deltas according to the
flowchart of FIG. 5 gives rise to 11 different strength modifiers 60 identified
as F1–F11.” Id. at 6:46–48. Figure 5 of Wilkins is reproduced below:
IPR2021-00827
Patent 9,781,448 B2
26

Wilkins’s Figure 5 is a flow chart diagram of a method for selecting a filter
strength modifier. Id. at 2:30–31. Petitioner points to no disclosure in
Wilkins of any of the eleven filter strength modifiers F1–F11 completely
eliminating baseline filter strength f. Thus, Petitioner’s augmenting of
Moon’s Table 1 by adding a Bs value of “0” (indicating no filtering) to the
table to correspond to non-split mode (Pet. 39) is lacking in rational
underpinning and evidently based on inappropriate hindsight reasoning.
9. Conclusion
For the foregoing reasons, Petitioner has not shown a reasonable
likelihood that it would prevail in establishing unpatentability of either claim
1 or claim 3 as obvious over Moon, Teng, and Wilkins.

IPR2021-00827
Patent 9,781,448 B2
27
D. Alleged Obviousness of
Claims 1 and 3 over Moon and Teng
1. Petitioner’s Combination of Moon and Teng
In this alleged ground of unpatentability, Petitioner foregoes all
reliance on Wilkins and relies solely on Moon and Teng. In so doing,
Petitioner no longer attempts to meet the “based on . . . [a or the] size of the
current block” recitation of claims 1 and 3 and addresses only the “based on
. . . [an or the] intra prediction mode . . . of the current block” recitation.
Pet. 63. That is because only one of the two recitations need to be met,
given the “at least one” modifier preceding the two recitations. Ex. 1001,
12:4–6, 12:21–22, 12:32–34, 12:49–50. Further, Petitioner no longer relies
on Wilkins, but on Teng, to add a loop filter to Moon. Id.
2. Deficiencies of Petitioner’s Application
of Moon and Teng to Claims 1 and 3
Because Petitioner no longer relies on Wilkins in this alleged ground
of unpatentability, the deficiencies noted above in Section II.C. in the
proposed combination of Moon, Teng, and Wilkins, stemming from Wilkins,
no longer exist. However, two deficiencies concerning Moon are equally
applicable here. First, as explained above in Section II.C.6., Moon’s filter
120 does not filter any prediction value as is required by both claims 1 and
3. Second, as explained above in Section II.C.7, Moon does not determine
whether to apply a filter based on a prediction mode of the current block as
is required by both claims 1 and 3 (given that Petitioner does not assert the
alternative condition, i.e., based on a size of the current block).

IPR2021-00827
Patent 9,781,448 B2
28
3. Conclusion
For the foregoing reasons, Petitioner has not shown a reasonable
likelihood that it would prevail in establishing unpatentability of either claim
1 or claim 3 as obvious over Moon and Teng.
III. CONCLUSION
Petitioner has not shown a reasonable likelihood that it would prevail
in establishing unpatentability of any challenged claim of the ’448 patent.
IV. ORDER
It is
ORDERED that the Petition is denied, and no inter partes review is
instituted on any claim over any alleged ground of unpatentability.

IPR2021-00827
Patent 9,781,448 B2
29
For PETITIONER:
Eric A. Buresh
Robin Snader
ERISE IP, P.A.
eric.buresh@eriseip.com
robin.snader@eriseip.com

Roshan S. Mansinghani
Ashraf A. Fawzy
Jung S. Hahm
Unified Patents LLC
roshan@unifiedpatents.com
afawzy@unifiedpatents.com
jung@unifiedpatents.com

For PATENT OWNER:
W. Karl Renner
Roberto J. Devoto
Nicholas Stephens
Fish & Richardson P.C.
axf-ptab@fr.com
devoto@fr.com
nstephens@fr.com

Kayvan B. Noroozi
NOROOZI PC
kayvan@noroozipc.com