`
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`
`____________________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`____________________
`
`UNIFIED PATENTS, LLC
`Petitioner
`
`v.
`
`ELECTRONICS AND TELECOMMUNICATIONS RESEARCH INSTITUTE,
`KWANGWOON UNIVERSITY RESEARCH INSTITUTE FOR INDUSTRY
`COOPERATION, INDUSTRY-ACADEMIA COOPERATION GROUP OF
`SEJONG UNIVERSITY,
`Patent Owners
`
`____________________
`
`IPR2021-00368
`Patent No. 9,736,484
`____________________
`
`
`
`PATENT OWNER’S PRELIMINARY RESPONSE
`PURSUANT TO 37 C.F.R. §42.107(a)
`
`
`
`
`
`
`
`United Patent, LLC. Ex. 1049 Page 1 of 78
`United Patents, LLC v. Elects. & Telecomm. Res. Inst., et al.
`IPR2021-00368
`
`
`
`IPR2021-00368
`U.S. Patent No. 9,736,484
`
`
`
`
`Patent Owners’ Preliminary Response
`
`TABLE OF CONTENTS
`
`
`I. SUMMARY OF ARGUMENT ......................................................................... 1
`II. CLAIM CONSTRUCTION ........................................................................... 2
`A. “horizontal intra prediction mode” / “vertical intra prediction mode” ... 4
`III. A PERSON OF ORDINARY SKILL IN THE ART ................................. 10
`IV. ARGUMENT ................................................................................................ 10
`A. Grounds 1 and 2: Prior Art Fails to Disclose Claimed Vertical and
`Horizontal Intra Prediction Modes ................................................................... 10
`1. Nishi’s AC Prediction in the vertical direction is not a vertical intra
`prediction mode ................................................................................................. 11
`2. Nishi’s AC Prediction in the horizontal direction is not a horizontal intra
`prediction mode ................................................................................................. 15
`B. Grounds 3 and 4: Prior Art Fails to Render Obvious Selecting a
`Horizontal or Vertical Intra Scanning Mode in Response to a Vertical or
`Horizontal Intra Prediction Mode .................................................................... 18
`1. Petition Confirms that Adaptive Scanning According to Do is Limited to
`Inter Prediction .................................................................................................. 21
`(1a) None of the applied references disclose selecting a scanning order
`based on directional intra prediction modes .................................................. 22
`(1b) Inter prediction is critical to Do’s invention and adaptive scanning
`patterns ........................................................................................................... 22
`(1c) Petitioner provides no evidence that Do’s inventive idea is operable
`with intra prediction ....................................................................................... 24
`(1d) Petitioner’s evidence indicates that Do’s inventive idea is NOT
`applicable to intra prediction ......................................................................... 26
`2. Petitioner Fails to Establish Prima Facie Case of Obviousness in
`Modifying Do in View of Kobayashi or Kalevo ............................................... 30
`(2a) Petitioner does not properly set forth difference between Challenged
`Claim and prior art ........................................................................................ 32
`
`
`
`i
`
`United Patent, LLC. Ex. 1049 Page 2 of 78
`United Patents, LLC v. Elects. & Telecomm. Res. Inst., et al.
`IPR2021-00368
`
`
`
`IPR2021-00368
`U.S. Patent No. 9,736,484
`
`Patent Owners’ Preliminary Response
`
`(2b) Petitioner does notprovide “afinding that the substituted components
`andtheirfunctions were known in the art?”.....2...2:.c:.cc1cc:00cseceeeeeeeeeeeeeeseeees34
`
`(2c) Petitioner does not establish that a POSITA could have substituted
`intra prediction modesfor contourfeatures to select a scanning pattern .....36
`
`(2d) Petition does not establish that claimed invention would have been a
`predictable resultfrom the proposed substitution. .........22..2:..10c1cc00eee0eeeeeeeees44
`
`V. THE PETITION SHOULD BE DISMISSED FOR FAILING TO NAME
`
`ALL REAL PARTIES IN INTEREST 0200002000000. 200 cc0 cece ccccceccceesceeseeeeeeeeeeeeeeeees 49
`
`A. Legal Standard... cocci cece cecceccccceecceeeeeeeeceeeeeecesceeceeseeeecesseeeeesteeeeeeees 51
`
`B. The Board Should Exercise its Authority to Determine RPIs to this
`Proceeding -..................ecceeeeeec cece cee eec cece ceeeeececeeecesseeeceeeceeseseeceeseseeseeeeeeeeeeeteeeeteeees 52
`Cc. Pe is an Unnamed RPIto this Proceeding............................56
`1. Relevant factual findings..............0 ccc ceeecceeeceeeeeeeeseeeceeeceeeeeeeecesseeeesseeeses 56
`2. BEshould be identified as an RPI ..............ecceececeececcesceeseceeeeeeeeeeeeeeees68
`
`VI. CONCLUSION022.2. ccc ccc ccccc cece cceccceceeece cece cececeseceseeesseceseeeseeesseesseeseeeesseeees 70
`
`il
`
`United Patent, LLC. Ex. 1049 Page 3 of 78
`United Patents. LLC v. Elects. & Telecomm.Res.Inst., et al.
`IPR2021-00368
`
`
`
`IPR2021-00368
`U.S. Patent No. 9,736,484
`
`
`TABLE OF AUTHORITIES
`
`
`
`Patent Owners’ Preliminary Response
`
`Cases
`Adaptics Limited v. Perfect Company LLC, IPR2018-01596,
`Institution Decision, Paper 20 (March 6, 2019) (precedential) ..................... 33, 34
`Ancora Techs., Inc. v. Apple, Inc., 744 F.3d 732, (Fed. Cir. 2014) ........................... 6
`Applications in Internet Time, LLC v. RPX Corp.,
`897 F.3d 1336 (Fed. Cir. 2018) ......................................................... 51, 52, 59, 67
`Ecolochem, Inc. v. S. Cal. Edison Co.,
`227 F.3d 1361 (Fed. Cir. 2000) ............................................................................ 29
`Eli Lilly & Co. v. Actavis Elizabeth LLC,
`435 F. App'x 917 (Fed. Cir. 2011) ........................................................................ 37
`Fall Line Patents, LLC v. Unified Patents, LLC,
`818 F.App’x 1014 (Fed. Cir. 2020) ............................................................... 52, 57
`Graham v. John Deere Co., 383 U.S. 1 (1966) ................................................ 32, 34
`Home Diagnostics, Inc. v. LifeScan, Inc., 381 F.3d 1352 (Fed. Cir. 2004)............... 3
`In re Dembiczak, 175 F.3d 994 (Fed. Cir. 1999) ..................................................... 29
`In re Gordon, 733 F.2d 900 (Fed. Cir. 1984) .......................................................... 26
`In re Hedges, 783 F.2d 1038 (Fed. Cir. 1986) .................................................. 30, 47
`In re Ratti, 270 F.2d 810 (C.C.P.A. 1959) ........................................................ 20, 26
`In re Wesslau, 353 F.2d 238 (CCPA 1965) ...................................................... 44, 46
`Intuitive Surgical, Inc., v. Ethicon Endo-Surgery, Inc.,
`IPR2018-00938, Institution Decision, Paper 8 ..................................................... 41
`KSR Int’l Co. v. Teleflex Inc., 550 U.S. 398 (2007) ......................................... 30, 38
`Microsoft Corporation et al. v. Koninklijke Philips N.V.,
`IPR2018-00185, Institution Decision, Paper No. 7, (May 22, 2018) ................... 20
`Phillips v. AWH Corp., 415 F. 3d 1303 (Fed. Cir. 2005) .......................................... 3
`Plas-Pak Indus. v. Sulzer Mixpac AG, 600 F. App'x 755 (Fed. Cir. 2015) ............. 26
`R.J. Reynolds Vapor Company v. Fontem Holdings 1 B.V.,
`IPR2016-01692, Final Written Decision, Paper 45, (March 2, 2018) ........... 21, 31
`
`
`
`iii
`
`United Patent, LLC. Ex. 1049 Page 4 of 78
`United Patents, LLC v. Elects. & Telecomm. Res. Inst., et al.
`IPR2021-00368
`
`
`
`
`
`Patent Owners’ Preliminary Response
`
`IPR2021-00368
`U.S. Patent No. 9,736,484
`
`RPX Corporation v. Applications in Internet Time, LLC,
`IPR2015-01750, Paper 128 (Oct. 2, 2020) (precedential) ............................ passim
`Saudi Arabian Oil Company v. SK Innovation Co., Ltd.,
`IPR2018-00159, Final Written Decision, Paper 21 .............................................. 35
`SharkNinja Operating LLC v. iRobot Corp.,
`IPR2020-00734, Paper 11 (Oct. 6, 2020) (precedential) ...................................... 52
`Standard Oil Co. v. Am. Cyanamid Co., 774 F.2d 448 (Fed. Cir. 1985) ......... 36, 48
`Unified Patents, LLC v. ETRI et al., IPR2020-01048, Paper 21 ............................. 59
`
`Statutes
`35 U.S.C. § 312(a) ...................................................................................... 50, 51, 70
`
`Regulations
`37 C.F.R. § 42.100(b) ......................................................................................... 2, 12
`
`
`
`
`
`
`
`
`iv
`
`United Patent, LLC. Ex. 1049 Page 5 of 78
`United Patents, LLC v. Elects. & Telecomm. Res. Inst., et al.
`IPR2021-00368
`
`
`
`United Patent, LLC. Ex. 1049 Page 6 of 78
`United Patents, LLC v. Elects. & Telecomm. Res. Inst., et al.
`IPR2021-00368
`
`
`
`IPR2021-00368
`U.S. Patent No. 9,736,484
`
`
`
`
`Patent Owners’ Preliminary Response
`
`Electronics and Telecommunications Research
`
`Institute
`
`(“ETRI”),
`
`Kwangwoon University Research
`
`Institute
`
`for
`
`Industry Cooperation
`
`(“Kwangwoon”), and Industry-Academia Cooperation Group of Sejong University
`
`(“Sejong”) (collectively, “Patent Owners”) submit this Preliminary Response to the
`
`Petition (“Pet.”) seeking inter partes review of claim 4 (“the Challenged Claim”) of
`
`U.S. Patent No. 9,736,484 (“the ‘484 Patent,” Ex-1001), filed by Unified Patents,
`
`LLC (“Unified” or “Petitioner”).
`
`I.
`
`SUMMARY OF ARGUMENT
`
`The Board should deny Unified’s Petition because it fails to establish a
`
`reasonable likelihood of success under any of Grounds 1-4.
`
`
`
`The ‘484 Patent discloses and claims an improved adaptive scanning
`
`technique to efficiently scan transform coefficients of an intra prediction difference
`
`image, i.e., the DCT-converted residual data indicating differences between pixel
`
`values of an original image block and a predicted block. Ex-1001, 5:26-32. The
`
`‘484 Patent’s inventors discovered that an optimal scanning mode could be selected
`
`in response to standardized directional intra prediction modes prevalent at that time.
`
`This is reflected in claim 1:
`
`select a horizontal scanning mode as the scanning mode when the
`intra prediction mode is a vertical intra prediction mode, and select
`a vertical scanning mode as the scanning mode when the intra
`prediction mode is a horizontal intra prediction mode.
`1
`
`
`
`United Patent, LLC. Ex. 1049 Page 7 of 78
`United Patents, LLC v. Elects. & Telecomm. Res. Inst., et al.
`IPR2021-00368
`
`
`
`IPR2021-00368
`U.S. Patent No. 9,736,484
`
`
`
`Conversely, none of the references applied by Petitioner suggests using
`
`
`
`Patent Owners’ Preliminary Response
`
`directional intra prediction modes to select scanning modes. Instead, Nishi (Grounds
`
`1, 2) discloses an adaptive scanning pattern selected based on a coefficient intra-
`
`frame prediction that a person of ordinary skill in the art (POSITA) would have
`
`understood as fundamentally different from the patent’s standardized intra
`
`prediction modes. Further, Do, Kobayashi, and Kalevo (Grounds 3, 4) all fail to
`
`suggest an adaptive scan based on an intra prediction mode. Instead, the contour
`
`feature-based adaptive scanning taught by Do would have led a POSITA away from
`
`the claimed invention. Indeed, Petitioner’s own evidence plainly shows that the
`
`prior art expressly taught the opposite of the claimed scanning mode selections when
`
`based on contour features.
`
`
`
`These and other deficiencies mean that the references applied in the Petition’s
`
`Grounds of unpatentability fall short of satisfying the elements recited by the
`
`Challenged Claim. Unified’s Petition should therefore be denied in its entirety.
`
`II. CLAIM CONSTRUCTION
`
`Terms of a claim subject to an inter partes review are “construed using the
`
`same claim construction standard that would be used to construe the claim in a civil
`
`action…, including construing the claim in accordance with the ordinary and
`
`customary meaning of such claims 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). Claim
`
`
`
`2
`
`United Patent, LLC. Ex. 1049 Page 8 of 78
`United Patents, LLC v. Elects. & Telecomm. Res. Inst., et al.
`IPR2021-00368
`
`
`
`IPR2021-00368
`U.S. Patent No. 9,736,484
`
`terms must be read “not only in the context of the particular claim in which the
`
`Patent Owners’ Preliminary Response
`
`
`
`disputed term appears, but in the context of the entire patent, including the
`
`specification.” Phillips v. AWH Corp., 415 F. 3d 1303, 1313 (Fed. Cir. 2005). There
`
`is a “heavy presumption” that claim terms “carry their accustomed meaning in the
`
`relevant community at the relevant time.” Home Diagnostics, Inc. v. LifeScan, Inc.,
`
`381 F.3d 1352, 1355 (Fed. Cir. 2004).
`
`
`
`Petitioner does not provide any claim construction analysis, stating instead
`
`that “claim terms in the Challenged Claims should receive their plain and ordinary
`
`meanings and that no express construction is needed to resolve the grounds presented
`
`herein.” Pet., 10. The interpretation of the claimed “vertical intra prediction mode”
`
`and “horizontal intra prediction mode,” however, is at issue in Petitioner’s
`
`application of references to the Challenged Claim. The “vertical intra prediction
`
`mode” and “horizontal intra prediction mode” in the ‘484 Patent are terms of art
`
`designating standardized spatial-domain directional prediction modes for predicting
`
`all pixels of a current block by using adjacent pixels. See, e.g., Ex-1001, 1:49-2:44,
`
`FIGS. 1-3 (standardized intra prediction modes); Ex-1002, ¶¶52-54. The applied
`
`Nishi reference (Ex-1014), meanwhile, implements an intra-frame prediction that is
`
`definitively not the standardized directional intra prediction modes of the ‘484
`
`Patent. Nishi’s intra-frame prediction is of only some (not all) AC coefficients (not
`
`pixels) of a current block in the transform (not spatial) domain by using distant (not
`
`
`
`3
`
`United Patent, LLC. Ex. 1049 Page 9 of 78
`United Patents, LLC v. Elects. & Telecomm. Res. Inst., et al.
`IPR2021-00368
`
`
`
`IPR2021-00368
`U.S. Patent No. 9,736,484
`
`adjacent) coefficients (not pixels). As the cited AC prediction of Nishi is
`
`Patent Owners’ Preliminary Response
`
`
`
`fundamentally different from the standardized directional prediction modes
`
`described in the ‘484 Patent and recognized in the art (confirmed by Petitioner’s own
`
`expert (Ex-1002 at ¶¶52-54)), Patent Owner submits that the construction of the
`
`claimed directional intra prediction modes will resolve Grounds 1 and 2.
`
`A. “horizontal intra prediction mode” / “vertical intra prediction mode”
`In the “TECHNICAL TUTORIAL” section of his Declaration (Ex-1002),
`
`Petitioner’s expert (Dr. Havlicek) characterizes “vertical” and “horizontal” intra
`
`prediction as “commonly used” intra prediction modes and expressly provides their
`
`plain and ordinary meanings in the art. Ex-1002 at ¶¶52-54. “For horizontal
`
`prediction, every pixel in the current block… is predicted by the pixel on the same
`
`line and immediately to the left of the current block.” Id., ¶52. “For vertical
`
`prediction, every pixel in the current block… is predicted by the pixel in the same
`
`column and immediately above the current block.” Id., ¶54.
`
`Patent Owner agrees. At the time of the invention, “horizontal intra prediction
`
`mode” and “vertical intra prediction mode” were standardized terms of art
`
`designating directional prediction modes for predicting “every pixel in the current
`
`block” by using “immediately” adjacent boundary pixels of neighboring blocks in
`
`the spatial domain. Ex-1002 at ¶¶49, 52-54.
`
`
`
`4
`
`United Patent, LLC. Ex. 1049 Page 10 of 78
`United Patents, LLC v. Elects. & Telecomm. Res. Inst., et al.
`IPR2021-00368
`
`
`
`IPR2021-00368
`U.S. Patent No. 9,736,484
`
`
`
`
`Patent Owners’ Preliminary Response
`
`Thus, the ordinary and customary meaning of “horizontal intra prediction
`
`mode” is, as Petitioner’s expert explains, a spatial prediction mode in which “every
`
`pixel in the current block… is predicted by the pixel on the same line and
`
`immediately to the left of the current block.” Id., ¶52. This can be understood from
`
`the below illustration provided by Dr. Havlicek, in which the current block is in
`
`orange and the neighboring blocks are grey:
`
`Id.
`
`
`
`Further, the ordinary meaning of “vertical intra prediction mode” is, as
`
`Petitioner’s expert explains, a spatial prediction mode in which “every pixel in the
`
`current block… is predicted by the pixel in the same column and immediately above
`
`
`
`5
`
`United Patent, LLC. Ex. 1049 Page 11 of 78
`United Patents, LLC v. Elects. & Telecomm. Res. Inst., et al.
`IPR2021-00368
`
`
`
`IPR2021-00368
`U.S. Patent No. 9,736,484
`
`the current block.” Id., ¶54. This can be understood from the below illustration
`
`Patent Owners’ Preliminary Response
`
`
`
`provided by Dr. Havlicek:
`
`Id.
`
`
`
`The standardized meanings of the directional intra prediction modes, as
`
`articulated by Dr. Havlicek, are consistent with the specification of the ‘484 Patent.
`
`To this end, the specification does not provide any special meanings for the
`
`directional intra prediction modes that depart from their standardized usages.
`
`Ancora Techs., Inc. v. Apple, Inc., 744 F.3d 732, 734 (Fed. Cir. 2014) ("A claim term
`
`should be given its ordinary meaning in the pertinent context, unless the patentee
`
`has made clear its adoption of a different definition or otherwise disclaimed that
`
`meaning”). Instead, the “horizontal” and “vertical” intra prediction modes of the
`
`
`
`6
`
`United Patent, LLC. Ex. 1049 Page 12 of 78
`United Patents, LLC v. Elects. & Telecomm. Res. Inst., et al.
`IPR2021-00368
`
`
`
`IPR2021-00368
`U.S. Patent No. 9,736,484
`
`‘484 Patent are in accord with their “standard” usages. Ex-1001, 1:64-2:44. This is
`
`Patent Owners’ Preliminary Response
`
`
`
`confirmed by Dr. Havlicek himself:
`
`Consistent with the understanding of the POSA, and as explained
`above, the ‘484 Patent explains that ‘spatial prediction’ uses
`‘boundary pixels of already recovered adjacent blocks’ and takes
`advantage of the fact that pixels adjacent to a pixel to be predicted
`‘are highly likely to have similar values.’… As discussed above, this
`type of intra prediction is used in H.264.”
`Ex-1002, ¶¶73-74 (citing Ex-1001, 1:49-61).
`
`The patent—like Dr. Havlicek—describes and illustrates the horizontal intra
`
`prediction mode (using a 4x4 block as an example) as the standardized spatial
`
`prediction mode in which all pixels of the current block are predicted using
`
`immediately adjacent boundary pixels of a left neighboring block. Ex-1001, 1:49-
`
`55, 2:35-44, FIG. 3; Ex-1002, ¶¶74-75. Referring to FIG. 3, reproduced below,
`
`“pixel a 205, pixel b 206, pixel c 207, and pixel d 208 are predicted based on an
`
`adjacent pixel I in a horizontal direction.” Ex-1001, 2:37-39; see also Ex-1002, ¶75.
`
`Similarly, “pixels e, f, g, and h are predicted based on an adjacent pixel J in the
`
`horizontal direction, and pixels i, j, k and l are predicted based on an adjacent pixel
`
`K in the horizontal direction. Pixels m, n, o and p are predicted based on an adjacent
`
`pixel L in the horizontal direction.” Ex-1001, 2:40-44; see also Ex-1002, ¶75. Thus,
`
`
`
`7
`
`United Patent, LLC. Ex. 1049 Page 13 of 78
`United Patents, LLC v. Elects. & Telecomm. Res. Inst., et al.
`IPR2021-00368
`
`
`
`IPR2021-00368
`U.S. Patent No. 9,736,484
`
`all pixels a-p of the current block are predicted in the spatial (or pixel) domain using
`
`Patent Owners’ Preliminary Response
`
`
`
`immediately adjacent boundary pixels I-L in the left neighboring block.
`
`Ex-1001, FIG. 3 (horizontal intra prediction mode)
`
`
`
`Further, the patent describes and illustrates the vertical intra prediction mode
`
`(using a 4x4 block as an example) as the standardized spatial prediction mode in
`
`which all pixels of the current block are predicted using immediately adjacent
`
`boundary pixels of a top neighboring block. Ex-1001, 1:49-55, 2:24-34, FIG. 2; see
`
`also Ex-1002, ¶74. As can be seen in FIG. 2, reproduced below, “pixel a 201, pixel
`
`e, 202, pixel i 203, and pixel m 204 are predicted based on an adjacent pixel D in the
`
`vertical direction.” Id., 2:26-28; see also Ex-1002, ¶74. “Also, pixels b, f, j and b
`
`are predicted based on an adjacent pixel B in the vertical direction, and pixels c, g,
`
`
`
`8
`
`United Patent, LLC. Ex. 1049 Page 14 of 78
`United Patents, LLC v. Elects. & Telecomm. Res. Inst., et al.
`IPR2021-00368
`
`
`
`IPR2021-00368
`U.S. Patent No. 9,736,484
`
`k and o are predicted based on an adjacent pixel C in the vertical direction. Pixels
`
`Patent Owners’ Preliminary Response
`
`
`
`d, h, l and p are predicted based on an adjacent pixel D in the vertical direction.” Id.,
`
`2:29-33; see also Ex-1002, ¶74. In other words, all pixels a-p of the current block
`
`are predicted in the spatial (or pixel) domain using immediately adjacent boundary
`
`pixels A-D in the top neighboring block.
`
`Id., FIG. 2 (vertical intra prediction mode)
`
`
`
`Therefore, as properly construed, the plain and ordinary meaning of the
`
`claimed “horizontal intra prediction mode” at the time of the invention is a spatial
`
`prediction mode in which “every pixel in the current block… is predicted by the
`
`pixel on the same line and immediately to the left of the current block.” Ex-1002,
`
`¶52. Similarly, the plain and ordinary meaning of the claimed “vertical intra
`
`prediction mode” at the time of the invention is a spatial prediction mode in which
`
`
`
`9
`
`United Patent, LLC. Ex. 1049 Page 15 of 78
`United Patents, LLC v. Elects. & Telecomm. Res. Inst., et al.
`IPR2021-00368
`
`
`
`IPR2021-00368
`U.S. Patent No. 9,736,484
`
`“every pixel in the current block… is predicted by the pixel in the same column and
`
`Patent Owners’ Preliminary Response
`
`
`
`immediately above the current block.” Id., ¶54.
`
`III. A PERSON OF ORDINARY SKILL IN THE ART
`
`PO agrees with Unified’s definition of a POSITA. Pet., 10.
`
`IV. ARGUMENT
`
`The Petition should be denied at least because the applied references do not
`
`anticipate or render obvious “selecting a horizontal scanning mode in response to
`
`the intra prediction mode being a vertical intra prediction mode” and “selecting a
`
`vertical scanning mode in response to the intra prediction mode being a horizontal
`
`intra prediction mode.”
`
`A. Grounds 1 and 2: Prior Art Fails to Disclose Claimed Vertical and
`Horizontal Intra Prediction Modes
`Grounds 1 and 2 of the Petition must fail because Nishi does not disclose
`
`directional intra prediction modes. Instead, Nishi discloses an AC prediction that is
`
`starkly different from the standardized vertical and horizontal intra prediction modes
`
`of the ‘484 Patent.
`
`As Petitioner’s expert, Dr. Havlicek, explains, the “horizontal intra prediction
`
`mode” is a spatial prediction mode in which “every pixel in the current block… is
`
`predicted by the pixel on the same line and immediately to the left of the current
`
`block.” Ex-1002, ¶52. Similarly, the “vertical intra prediction mode” is a spatial
`
`
`
`10
`
`United Patent, LLC. Ex. 1049 Page 16 of 78
`United Patents, LLC v. Elects. & Telecomm. Res. Inst., et al.
`IPR2021-00368
`
`
`
`IPR2021-00368
`U.S. Patent No. 9,736,484
`
`prediction mode in which “every pixel in the current block… is predicted by the
`
`Patent Owners’ Preliminary Response
`
`
`
`pixel in the same column and immediately above the current block.” Id., ¶54.
`
`Conversely, the AC prediction implemented in Nishi is a transform coefficient
`
`(NOT spatial) prediction mode in which AC coefficients (NOT pixels) of only one
`
`(NOT every) row or column is predicted by a transform coefficient (NOT pixel) of a
`
`row or a column on a distant (NOT immediately adjacent) boundary of a neighboring
`
`block. Unlike the “vertical intra prediction mode” and “horizontal intra prediction
`
`mode” of the ‘484 Patent, Nishi’s AC prediction of transform coefficients does not
`
`and cannot “take[] advantage of a characteristic that when a pixel is predicted, pixels
`
`adjacent to it are highly likely to have similar values.” Ex-1001, 52-55.
`
`1. Nishi’s AC Prediction in the vertical direction is not a vertical intra
`prediction mode
`The Petition relies on Nishi’s AC intra-frame prediction in a vertical direction
`
`for an alleged teaching of the claimed vertical intra prediction mode. Pet., 31-33,
`
`35. Petitioner repeatedly refers to this AC prediction as a “vertical intra prediction
`
`mode” (id.), but provides no objective evidence that a POSITA would have
`
`considered the AC prediction to be a “vertical intra prediction mode” per its
`
`customary meaning at the time of the invention. For example, Nishi itself never
`
`characterizes the AC prediction as a “vertical intra prediction mode”; this descriptor
`
`originates from Petitioner without explanation or evidence. For at least this reason,
`
`
`
`11
`
`United Patent, LLC. Ex. 1049 Page 17 of 78
`United Patents, LLC v. Elects. & Telecomm. Res. Inst., et al.
`IPR2021-00368
`
`
`
`IPR2021-00368
`U.S. Patent No. 9,736,484
`
`Grounds 1 and 2 fail to establish that a POSITA would have understood Nishi’s
`
`Patent Owners’ Preliminary Response
`
`
`
`compression technique to correspond to the claimed vertical intra prediction mode.
`
`Further, the ‘484 Patent and the Petition both distinguish the claimed
`
`directional intra prediction modes standardized at the time of the invention from the
`
`prior art (MPEG-2) transform coefficient (DCT)-based intra-frame prediction
`
`described in Nishi. Ex-1014, 1:18, 1:24, 2:24 (MPEG). For example, the ‘484 Patent
`
`explains that the H.264 standard is able to achieve much higher compression rates
`
`than the prior MPEG standards “by using such technique as intra prediction
`
`encoding.” Ex-1001, 1:56-65; see also Ex-1017, 1 (H.264 compression efficiency
`
`higher than MPEG-4, H.263, etc., because of “new advanced coding tools such as…
`
`spatial prediction for intra coding”). The Petition, too, distinguishes the directional
`
`intra prediction modes standardized at the time of the ‘484 Patent from the prior
`
`video compression techniques. Pet., 40; Ex-1002, ¶184.
`
`In either case, by 2006 (the time of the invention), the term “vertical intra
`
`prediction mode” had a well-understood and standardized meaning in the art. Ex-
`
`1002, ¶¶49, 54, 73-74. Pursuant to its “ordinary and customary meaning” (37 C.F.R.
`
`§42.100(b)), a “vertical intra prediction mode” is a spatial prediction mode in which
`
`“every pixel in the current block… is predicted by the pixel in the same column and
`
`immediately above the current block.” Ex-1002, ¶54; see supra §II(A). Nishi’s AC
`
`prediction in the vertical direction, meanwhile, is not a spatial prediction mode, does
`
`
`
`12
`
`United Patent, LLC. Ex. 1049 Page 18 of 78
`United Patents, LLC v. Elects. & Telecomm. Res. Inst., et al.
`IPR2021-00368
`
`
`
`United Patent, LLC. Ex. 1049 Page 19 of 78
`United Patents, LLC v. Elects. & Telecomm. Res. Inst., et al.
`IPR2021-00368
`
`
`
`IPR2021-00368
`U.S. Patent No. 9,736,484
`
`
`
`
`Patent Owners’ Preliminary Response
`
`First, the AC prediction in Nishi is not of pixels, but of quantized DCT
`
`coefficients. Ex-1014, 4:1-6. DCT coefficients are not synonymous with pixels, but
`
`are the result of conversion from the pixel/spatial domain to the transform/frequency
`
`domain. Ex-1002, ¶¶62-64. Each coordinate in the DCT-converted image block
`
`does not map to a corresponding pixel value; rather, each coordinate corresponds to
`
`a different frequency coefficient representing different directional frequencies (or
`
`“amplitudes of cosine waves of increasing frequency”) of the image data. Id., ¶63
`
`(quoting Ex-1022, 2:13-45). The spatial relationships between pixels are not
`
`embodied by the DCT coefficients. As such, the AC prediction is not a spatial
`
`prediction mode that predicts pixels in the spatial domain in order “to reduce spatial
`
`redundancies based on the notion that, in most cases, pixels that are near one
`
`another… have similar values.” Ex-1002, ¶49; see also Ex-1001, 1:49-55. Rather,
`
`Nishi discloses the conventional transform or frequency prediction mode for
`
`predicting frequency components in the frequency domain:
`
`The two-dimensional discrete cosine transform is a pair of
`mathematical equations that transforms one NxN array of numbers
`to or from another NxN array of numbers. The first array typically
`represents an NxN array of spatially determined pixel values which
`form the digital image. The second array is an array of discrete
`cosine transform coefficients which represent the image in the
`frequency domain.
`Ex-1022, 1:13-19 (quoted by Ex-1002, ¶63).
`14
`
`
`
`United Patent, LLC. Ex. 1049 Page 20 of 78
`United Patents, LLC v. Elects. & Telecomm. Res. Inst., et al.
`IPR2021-00368
`
`
`
`IPR2021-00368
`U.S. Patent No. 9,736,484
`
`
`
`
`Patent Owners’ Preliminary Response
`
`Second, the AC prediction in Nishi does not predict all coefficients (let alone
`
`“every pixel,” Ex-1002, ¶54) in the current block. Instead, the conventional AC
`
`prediction predicts only a single row—“the uppermost line”—of quantized
`
`transform coefficients. Ex-1014, 4:1-6.
`
`Third, the AC prediction does not predict coefficients (let alone pixels) in the
`
`current block using coefficients (let alone pixels) “immediately above the current
`
`block” (Ex-1002, ¶54). Instead, the reference row of coefficients in Nishi is the
`
`distant top row of the neighboring block. Where an 8x8 DCT sub-block is predicted,
`
`as in Nishi’s FIG. 30, the reference pixels are eight rows above the current block.
`
`In sum, the AC prediction in the vertical direction disclosed by Nishi is not a
`
`spatial prediction mode in which “every pixel in the current block… is predicted by
`
`the pixel in the same column and immediately above the current block.” Ex-1002,
`
`¶54. Nishi therefore does not disclose or render obvious1 a “vertical intra prediction
`
`mode” as recited in the Challenged Claim.
`
`2. Nishi’s AC Prediction in the horizontal direction is not a horizontal intra
`prediction mode
`Nishi also does not disclose a “horizontal intra prediction mode” for all the
`
`same reasons. For this standardized directional mode, the Petition relies on Nishi’s
`
`
`1 Ground 2 based on Nishi is raised only to allege obviousness of software
`
`implementation. Pet., 27-28.
`
`
`
`15
`
`United Patent, LLC. Ex. 1049 Page 21 of 78
`United Patents, LLC v. Elects. & Telecomm. Res. Inst., et al.
`IPR2021-00368
`
`
`
`IPR2021-00368
`U.S. Patent No. 9,736,484
`
`AC prediction in a horizontal direction. The horizontal AC prediction is not of
`
`Patent Owners’ Preliminary Response
`
`
`
`pixels, but of quantized DCT coefficients. Ex-1014, 4:6-11. Further, the horizontal
`
`AC prediction does not predict all coefficients (let alone “every pixel,” Ex-1002,
`
`¶52) in the current block (X), but only a single column—“the leftmost line”—of
`
`quantized transform coefficients. Ex-1014, 4:6-11. The conventional AC prediction
`
`also fails to predict coefficients (let alone pixels) in the current block (X) using
`
`coefficients (let alone pixels) “immediately to the left of the current block” (Ex-
`
`1002, ¶52). Instead, the reference column of coefficients in Nishi is the distant left-
`
`most row of the neighboring block, eight columns to the left of the current block:
`
`
`
`
`
`
`
`16
`
`United Patent, LLC. Ex. 1049 Page 22 of 78
`United Patents, LLC v. Elects. & Telecomm. Res. Inst., et al.
`IPR2021-00368
`
`
`
`United Patent, LLC. Ex. 1049 Page 23 of 78
`United Patents, LLC v. Elects. & Telecomm. Res. Inst., et al.
`IPR2021-00368
`
`
`
`IPR2021-00368
`U.S. Patent No. 9,736,484
`
`therefore does not disclose or render obvious the standardized “horizontal intra
`
`Patent Owners’ Preliminary Response
`
`
`
`prediction mode” recited in the Challenged Claim.
`
`Petitioner characterizes, without support, Nishi’s AC prediction as a “vertical
`
`intra prediction mode” and a “horizontal intra prediction mode,” but fails to establish
`
`that a POSITA would have the same understanding. Instead, Petitioner provides an
`
`expert Declaration that confirms the customary usage of the Challenged Claim’s
`
`standardized directional intra prediction modes. Petitioner fails to address the
`
`glaring differences between the standardized intra prediction modes, as recognized
`
`by Dr. Havlicek, and Nishi’s prediction technique. Grounds 1 and 2, therefore, must
`
`fail as Petitioner has not met its burden in establishing a reasonable likelihood that
`
`Nishi discloses or renders obvious e