`
`UNITED STATES INTERNATIONAL TRADE COMMISSION
`WASHINGTON, D.C.
`
`Before the Honorable Doris Johnson Hines
`Administrative Law Judge
`
`In the Matter of:
`
`CERTAIN VIDEO CAPABLE DEVICES,
`INCLUDING COMPUTERS,
`STREAMING DEVICES, TELEVISIONS,
`CAMERAS, AND COMPONENTS AND
`MODULES THEREOF
`
`
`
`
`
`
`
`
`
`
`Investigation No. 337-TA-1379
`
`COMPLAINANTS’ REPLY CLAIM CONSTRUCTION BRIEF
`
`
`
`AMAZON-1020
`7,532,808
`
`

`

`2.
`
`3.
`
`INTRODUCTION ...............................................................................................................1
`CLARIFICATION OF THE ’808 PATENT AND TECHNICAL BACKGROUND
`..............................................................................................................................................3
`1.
`The ’808 Patent’s Starting Point for the Redefined Skip Coding
`Mode ............................................................................................................4
`a.
`The three prior art approaches discussed by the ’808 Patent ...........5
`b.
`The ’808 Patent’s starting point for lexicography ...........................6
`Coding Modes and Residual Information ....................................................7
`a.
`Coding modes ..................................................................................8
`b.
`Residual information ......................................................................10
`Whether Motion Information is Transmitted in or Derived From the
`Redefined Skip Coding Mode ....................................................................12
`ARGUMENT .....................................................................................................................15
`A.
`Respondents Misapply Claim Differentiation .......................................................15
`B.
`Respondents Fail to Show a Single Embodiment Where Residual
`Information is Provided with Skip Coding Mode ..................................................17
`Nokia and the Staff are Following Lexicography, not Improperly Elevating
`a Single Embodiment .............................................................................................20
`a.
`Even if Nokia and the Staff are “improperly elevating a
`single embodiment” (and they are not), Respondents’
`construction still fails .....................................................................22
`CONCLUSION ..................................................................................................................23
`
`
`I.
`II.
`
`III.
`
`IV.
`
`
`
`C.
`
`TABLE OF CONTENTS
`
`
`
`i
`
`INV. NO. 337-TA-1379
`COMPLAINANTS’ REPLY CLAIM CONSTRUCTION BRIEF
`
`

`

`TABLE OF AUTHORITIES
`
`
`
`Page(s)
`
`Cases
`
`Enzo Biochem Inc. v. Applera Corp.,
`780 F.3d 1149 (Fed. Cir. 2015)................................................................................................16
`
`Honeywell Int’l, Inc. v. Universal Avionics Sys. Corp.,
`493 F.3d 1358 (Fed. Cir. 2007)................................................................................................15
`
`In re Katz Interactive Call Processing Pat. Litig.,
`639 F.3d 1303 (Fed. Cir. 2011)................................................................................................22
`
`Littelfuse, Inc. v. Mersen USA EP Corp.,
`29 F.4th 1376 (Fed. Cir. 2022) ..........................................................................................22, 23
`
`Multiform Desiccants, Inc. v. Medzam, Ltd.,
`133 F.3d 1473 (Fed. Cir. 1998)................................................................................................16
`
`Nystrom v. TREX Co.,
`424 F.3d 1136 (Fed. Cir. 2005)..........................................................................................16, 17
`
`Phillips v. AWH Corp.,
`415 F.3d 1303 (Fed. Cir. 2005) (en banc) ....................................................................... passim
`
`Tandon Corp. v. U.S. Int’l Trade Comm’n,
`831 F.2d 1017 (Fed. Cir. 1987)................................................................................................17
`
`Wi-LAN USA, Inc. v. Apple Inc.,
`830 F.3d 1374 (Fed. Cir. 2016)................................................................................................17
`
`Statutes
`
`35 U.S.C. ¶ 112 ..............................................................................................................................12
`
`
`
`
`
`
`ii
`
`INV. NO. 337-TA-1379
`COMPLAINANTS’ REPLY CLAIM CONSTRUCTION BRIEF
`
`

`

`I.
`
`INTRODUCTION
`
`This is a debate over whether the ’808 Patent’s definition of skip coding mode should be
`
`amended to include “and residual information can be provided for each skip mode macroblock.”
`
`It should not. If the inventor wanted to include “residual information” in their lexicography of skip
`
`coding mode, they would have done so.
`
`Nokia and the Staff support a construction copied directly from the inventor’s lexicography
`
`that is supported without contradiction throughout the patent. The inventor intentionally did not
`
`include “residual information” in the patent’s redefinition of “skip coding mode.” Black letter
`
`claim construction law tells us that the inventor’s chosen lexicography is dispositive.
`
`Respondents seek to alter the inventor’s lexicography despite the Federal Circuit’s
`
`directive in Phillips that “the inventor’s lexicography governs,” despite the mountain of evidence
`
`in the intrinsic record contradicting their construction, and despite Dr. Michael Orchard’s expert
`
`declaration that a person of ordinary skill in the art (“POSITA”) would see Respondents’
`
`construction as oxymoronic. (Amazon’s expert, Dr. Joseph Havlicek, does not affirmatively agree
`
`with Respondents’ construction in his declaration.) Respondents submit three arguments to
`
`overcome these hurdles. None succeeds.
`
`One, Respondents advance a claim differentiation theory and allege Nokia and the Staff’s
`
`construction invalidates four dependent claims. As a threshold matter, Respondents’ application
`
`of claim differentiation is mistaken. Respondents confuse a dependent limitation stating that no
`
`residual information is provided for a given macroblock in the encoded bitstream with the fact the
`
`redefinition of skip coding mode—just one part of what might be in a bitstream—does not include
`
`residual information. Of course, an entire encoded bitstream is a different concept than a block-
`
`1
`
`INV. NO. 337-TA-1379
`COMPLAINANTS’ REPLY CLAIM CONSTRUCTION BRIEF
`
`

`

`by-block coding mode. Respondents themselves implicitly admit this by stating residual
`
`information can be used in different contexts, not just in block-by-block coding.
`
`Moreover, we can assume arguendo that claim differentiation applies, and Respondents
`
`still lose. Lexicography and the intrinsic record govern over claim differentiation, which is only a
`
`guide rather than a rigid rule.
`
`Two, Respondents allege their construction is consistent “with all disclosed embodiments.”
`
`Respondents’ Brief (“Resp. Br.”) at 20. Respondents offer three so-called “embodiments.” The
`
`first so-called embodiment (“GMVC_16”) on the list is not even from the ’808 Patent. Id. (citing
`
`“Ex. 4 at 3” rather than anything in the patent). What Respondents describe as an embodiment is
`
`actually a prior art technique expressly rejected by the inventor as failing to sufficiently solve the
`
`problems in the art. See, e.g., 13:31-44 (background section identifying the problems with the prior
`
`art). The second and third so-called embodiments fare no better. Respondents identify Figures 6
`
`(encoder) and 7 (decoder) and advance the uncontroversial proposition that each figure is capable
`
`of non-skip coding modes while ignoring the repeated discussions in the patent that additional
`
`information, including residual and motion information, is not used by the redefined skip coding
`
`mode. The fact the patent describes an encoder and decoder capable of non-skip coding modes
`
`does not dictate the construction of skip coding mode. The ’808 Patent plainly states that non-skip
`
`coding modes include residual and motion information while the redefinition of skip coding mode
`
`does not include this additional information.
`
`Three, Respondents allege that Nokia and the Staff are elevating one embodiment in the
`
`patent over others. However, Respondents fail to identify a single embodiment of skip coding
`
`mode that includes residual information. Nothing in the patent suggests such an embodiment.
`
`Respondents ignore the ’808 Patent’s repeated descriptions of “the invention,” which all state that
`
`2
`
`INV. NO. 337-TA-1379
`COMPLAINANTS’ REPLY CLAIM CONSTRUCTION BRIEF
`
`

`

`no additional information is provided in the bitstream for skip coding mode. ’808 Patent at 14:62-
`
`64 (“according to the invention,” no additional information in the bitstream for skip coding mode);
`
`15:7-10 (same); 15:66-67 (same); 20:53-55 (same); 19:17-20 (same, albeit for an embodiment).
`
`A POSITA would readily understand that the intent of the ’808 Patent was to redefine the
`
`“SKIP” mode of Joint Model Number 1 or “JM1.” As Dr. Orchard explains in his declaration, a
`
`POSITA would readily understand from the lexicography and the specification that the “skip”
`
`1 That is, the ’808 Patent’s redefined
`aspect of the redefined skip coding mode remained intact.0F
`
`2 What
`skip coding mode—as its name suggests to POSITAs—skips residual information.1F
`
`Respondents are trying to do is take the “skip” out of “skip coding mode” by adding residual
`
`information to the definition of “skip coding mode.” This is despite the fact the inventor made the
`
`choice to not include “residual information” in the lexicography of “skip coding mode.” For the
`
`reasons below, this is legally and technically baseless. The ALJ should adopt Nokia and the Staff’s
`
`construction.
`
`II.
`
`CLARIFICATION OF THE ’808 PATENT AND TECHNICAL BACKGROUND
`
`Respondents’ brief suggests a misunderstanding of three basic concepts: (1) the ’808
`
`Patent’s JM1 starting point for its redefined skip coding mode; (2) coding modes and residual
`
`information; and (3) whether motion information is transmitted in or derived from the redefined
`
`skip coding mode. To set the stage for the argument that follows, these three concepts will be
`
`clarified in turn.
`
`
`1 Exhibit 2 to Complainants’ Opening Claim Construction Brief, Declaration of Dr. Michael
`Orchard (“Orchard Dec.”).
`2 Skip coding modes skip the representations of blocks in bitstreams and thus also skip motion
`vector information in the bitstream, but that is not the subject of this claim construction dispute.
`3
`
`INV. NO. 337-TA-1379
`COMPLAINANTS’ REPLY CLAIM CONSTRUCTION BRIEF
`
`

`

`1.
`
`The ’808 Patent’s Starting Point for the Redefined Skip Coding Mode
`
`Respondents represent that the ’808 Patent “alleges to improve upon ‘Global Motion
`
`Vector Coding’” and “purportedly changed how the prior art global-motion coding modes derived
`
`motion information.” Resp. Br. at 1 (citing ’808 Patent at 13:31-33) and 9 (citing ’808 Patent at
`
`Abstract). This misreads the patent.
`
`The ’808 Patent’s background section discusses three prior art approaches: (1) JM1; (2)
`
`global motion compensation or “GMC”; and (3) Global Motion Vector Coding or “GMVC”. E.g.,
`
`’808 Patent at 13:49-14:10. Respondents portray the ’808 Patent as building off of GMC or GMVC
`
`rather than JM1. See, e.g., Resp. Br. at 1, 9. This is perhaps an attempt to bake elements of GMC
`
`and GMVC into their proposed construction of the ’808 Patent’s “skip coding mode.”
`
`Respondents’ portrayal, however, is contradicted by the ’808 Patent.
`
`The ’808 Patent explains the starting point for its lexicography of “skip coding mode” was
`
`not GMC or GMVC but instead JM1. As the ’808 Patent expressly states, “[T]he present invention
`
`is based on a redefinition of the skip mode concept used in JM1 of the JVT codec.” Id. at 14:14-
`
`22; see also 17:43-53. The Respondents agree JM1 “uses no motion and no residual information.”
`
`Resp. Br. at 9.
`
`Moreover, the ’808 Patent explains that the very elements Respondents are trying to bake
`
`into the ’808 Patent’s skip coding mode (namely additional information, such as residual and
`
`motion information) are in fact differences between the present invention and GMC and GMVC:
`
`As will become apparent from the detailed description of the invention presented
`below, redefinition of the skip mode concept as proposed by the present invention
`has significant technical advantages compared with the previously described
`prior art video coding methods. In particular, the method according to the invention
`enables global and regional motion within a video sequence to be taken account of
`in an efficient manner without the need for complex warping of the reference frame
`or any other computationally demanding operations. Furthermore, in contrast to
`both the global motion compensation [GMC] and global motion vector coding
`4
`
`INV. NO. 337-TA-1379
`COMPLAINANTS’ REPLY CLAIM CONSTRUCTION BRIEF
`
`

`

`[GMVC] methods previously described, no additional information must be
`transmitted in the video bit-stream to enable correct decoding of the video data.
`Additionally, a minimal amount of modification is required to incorporate the
`method according to the invention into existing video coding systems that employ
`the concept of skip mode macroblocks.
`
` ’808 Patent at 14:65-15:14 (emphasis added).
`
`a.
`
`The three prior art approaches discussed by the ’808 Patent
`
`The inventions of the ’808 Patent arose in March 2002, when a video recommendation was
`
`under development. ’808 Patent at 10:23-39. As the ’808 Patent’s background section explains, a
`
`version of the recommendation as of January 2002 was described in a document called JM1. Id. at
`
`10:31-39. JM1 included a number of macroblock coding modes depicted in Table 3 of the ’808
`
`Patent. One of these coding modes was called “SKIP.” Id. at Table 3. In JM1, SKIP mode indicated
`
`that a macroblock “is to be copied from the reference video frame without using motion
`
`compensated prediction.” Id. at 10:63-67.
`
`JM1’s SKIP mode did not involve any information in the bitstream other than the indication
`
`of the coding mode. Orchard Dec. at ¶ 19; Exhibit 1 to Complainants’ Opening Claim Construction
`
`Brief (JM1) at Subclause 3.4.1 (page 17, NOK_AMHP_ITC00079650). The ’808 Patent describes
`
`that JM1 SKIP had the smallest overhead of the JM1 coding modes and JM1 assumed SKIP was
`
`“statistically the most likely coding mode.” ’808 Patent at 12:18-22, Table 3. However, JM1’s
`
`assumption was “not valid” because SKIP could not be used in video sequences containing
`
`“global” motions such as panning and zooming. Id. at 12:41-47. This hurt JM1’s compression
`
`efficiency because JM1’s higher overhead macroblock coding modes were required to represent
`
`this motion. Id.
`
`The ’808 Patent’s background section then explains attempts to solve this existing problem.
`
`Specifically, the patent discusses two “prior art solutions for improving the coding efficiency of
`
`5
`
`INV. NO. 337-TA-1379
`COMPLAINANTS’ REPLY CLAIM CONSTRUCTION BRIEF
`
`

`

`motion compensated prediction in the presence of global motion.” These solutions were called
`
`“global motion compensation” (GMC) and “Global Motion Vector Coding” (GMVC). ’808 Patent
`
`at 12:48-58. The ’808 Patent explains that each of the existing three approaches—JM1, GMC, and
`
`GMVC—“has some form of technical shortcoming.” Id. at 13:49-51. First, JM1 could not account
`
`for global motion and suffered efficiency degradation because any motion needed to be “explicitly
`
`model[ed].” Id. at 13:51-58. Second, GMC relied on a computationally complex reference frame
`
`“warping process” that required “additional information in the bit-stream to enable correct
`
`decoding of the video sequence.” Id. at 13:58-67. Third, GMVC, while computationally less
`
`demanding than GMC, nonetheless increased encoder complexity and “additional information
`
`must still be transmitted in the video bit-stream to enable correct decoding of the video data.” Id.
`
`at 13:67-14:5.
`
`As a result of the existing problems in the art, the inventor sought to “combine the
`
`simplicity of local motion compensation with the coding efficiency of global motion compensation
`
`to yield a video coding system with significantly improved compression performance and a
`
`negligible increase in complexity.” ’808 Patent at 14:6-10. Moreover, unlike the GMC and GMVC
`
`techniques, the invention not only provides an “improvement in coding efficiency in the presence
`
`of global motion (i.e. motion affecting the entire area of video frame), but also enables regional
`
`motion to be represented in an efficient manner.” Id. at 14:18-22.
`
`b.
`
`The ’808 Patent’s starting point for lexicography
`
`Respondents’ representations to the ALJ that the inventor sought to “improve upon” and
`
`“changed” GMC and GMVC (see Resp. Br. at 1, 9) are contradicted by the intrinsic record.
`
`The’808 Patent’s inventor rejected the GMC and GMVC approaches. E.g., ’808 Patent at 15:7-14,
`
`13:58-14:5. Instead, the inventor’s starting point was JM1. “[T]he present invention is based on a
`
`6
`
`INV. NO. 337-TA-1379
`COMPLAINANTS’ REPLY CLAIM CONSTRUCTION BRIEF
`
`

`

`redefinition of the skip mode concept used in JM1 of the JVT codec.” Id. at 15:13-18. The patent
`
`reiterates this point more than once. See id. at 17:43-53, 23:13-17. The inventor’s lexicography,
`
`which refers to JM1 (and not GMC or GMVC), is as follows:
`
`According to the invention, the coding modes used by encoder 600 correspond to
`those provided by JM1 of the JVT codec (shown in Table 3), with the exception
`that the SKIP mode is redefined to allow representation of global and regional
`motion. More specifically, the SKIP mode is modified in such a way that a zero
`(non-active) motion vector or a non-zero (active) motion vector is associated with
`each skip mode macroblock, depending on the characteristics of the motion in
`image segments surrounding the macroblock in question. In the following this type
`of motion vector will be referred to as a “skip mode motion vector”.
`
`’808 Patent at 17:43-53.
`
`
`The starting point for the inventor’s lexicography matters. As Dr. Orchard explains in his
`
`declaration, a POSITA would readily understand that JM1’s skip coding mode included no residual
`
`information and that the redefinition of skip coding mode likewise includes no residual
`
`information. Orchard Dec. at ¶ 19-20. Indeed, the inventor explicitly distinguished the invention
`
`from GMC and GMVC because the invention requires “no additional information.” ’808 Patent at
`
`14:65-15:14. There is no basis to add “residual information” to the ’808 Patent’s definition of skip
`
`coding mode simply because the patent explains why GMC and GMVC were inadequate solutions
`
`as background discussion. Indeed, the patent explains that GMC and GMVC are inadequate
`
`precisely because they require additional information.
`
`2.
`
`Coding Modes and Residual Information
`
`One point of agreement between the parties is that “prediction error,” a term used in the
`
`’808 Patent specification, can be called “residual information.” See Orchard Dec. at ¶ 18; Resp.
`
`Br. at 8. However, there appears to be some confusion about the ’808 Patent’s discussion of
`
`7
`
`INV. NO. 337-TA-1379
`COMPLAINANTS’ REPLY CLAIM CONSTRUCTION BRIEF
`
`

`

`
`
`prediction error and its relationship to coding modes.2F3 For example, Respondents state that
`
`“[r]esidual information can be used with global motion and object motion.” Resp. Br. at 8 (citing
`
`’808 Patent at 12:54-58, 3:31-35). Respondents rely on two disclosures from the ’808 Patent to
`
`support their statement: 12:54-58 and 3:31-35. But neither supports the proposition. The first
`
`citation, 12:54-58, does not say anything about prediction error or residual information. The
`
`second citation, 3:31-35, likewise does not say anything about “global motion” or “object motion.”
`
`Respondents’ claim differentiation argument is that certain dependent claims3F
`
`4 of the ’808
`
`Patent recite “wherein no residual information is provided for the first segment in the encoded
`
`bitstream” and therefore the independent claims “must allow providing residual information.”
`
`Resp. Br. at 18. However, whether residual information is included for a macroblock (a first
`
`segment) in the encoded bitstream at all is separate from whether that residual information is
`
`provided as part of skip coding mode for that macroblock. Where and how residual information
`
`can be included in a bitstream (e.g., by macroblock or for “global” or “object” motions), and
`
`whether the ’808 Patent gives any reason to believe its redefinition of skip coding mode includes
`
`residual information are therefore both relevant to this claim construction dispute.
`
`Clarification on coding modes and residual information is warranted.
`
`a.
`
`Coding modes
`
`The ’808 Patent describes that coding mode assignment is done on a macroblock-by-
`
`5 Encoder 600 (from Figure 6) acts “in a manner similar to the previously
`macroblock basis.4F
`
`
`3 INTER-coding modes, such as the NxM modes listed in Table 3 of the ’808 Patent, involve
`predictions made from a reference frame. See, e.g., ’808 Patent at 3:11-28. INTER-coding modes
`exploit temporal redundancy across frames. See, e.g., id. This is distinct from INTRA-coding
`modes, which exploit spatial redundancy within frames. See, e.g., id. at 2:56-10.
`4 Claims 30, 41, 49, and 60.
`5 A macroblock is a 16x16 block of pixels. ’808 Patent at 10:40-41.
`8
`
`INV. NO. 337-TA-1379
`COMPLAINANTS’ REPLY CLAIM CONSTRUCTION BRIEF
`
`

`

`described JVT codec” and assigns a coding mode to each macroblock “depending on the
`
`characteristics of the macroblock and the motion in the sequence being coded.” ’808 Patent at
`
`17:16-21. This process is called “motion estimation,” where an encoder will determine the best
`
`coding mode for a given macroblock based on the “overall cost value” (which may include, for
`
`example, an estimated number of bits to code the mode and represent motion vectors, if any). Id.
`
`at 17:21-42. Once the encoder performs “motion estimation operations for each of the available
`
`coding modes, including skip mode as modified according to the invention, encoder 600
`
`determines which coding mode yields the smallest overall cost value and selects that mode as the
`
`coding mode for the macroblock in question.” Id. at 18:12-17. The encoder then includes an
`
`indication of the selected coding mode for the macroblock in the bitstream. Id. at 18:17-20. This
`
`indication will enable a decoder to identify the coding mode for the macroblock and use the
`
`“correct form of motion-compensated prediction.” Id. at 18:20-23.
`
`The ’808 Patent describes that decoding can also occur on a macroblock-by-macroblock
`
`basis. ’808 Patent at 20:43-46. The ’808 Patent expressly states that what information is included
`
`in the bitstream for a given coding mode will depend on the coding mode:
`
`Depending on the coding mode, the compressed video data included in the bit-
`stream for an INTER-coded macroblock may comprise a combination of VLC
`encoded prediction error information for each block, motion vector information for
`the macroblock (or sub-blocks) and encoded control information including an
`indication of the coding mode used to encode the macroblock in question. If a
`macroblock is encoded in skip mode, no prediction error or motion vector
`information relating to the macroblock is included in the bit-stream.
`
`’808 Patent at 20:46-55 (emphasis added). In other words, the bitstream may include prediction
`
`error and motion vector information for non-skip coding modes. However, the bitstream does not
`
`include prediction error and motion vector information for skip coding mode.
`
`9
`
`INV. NO. 337-TA-1379
`COMPLAINANTS’ REPLY CLAIM CONSTRUCTION BRIEF
`
`

`

`
`
`This delineation between non-skip coding modes and skip coding mode is consistent with
`
`the patent’s statement that “According to the invention, the coding modes used by encoder 600
`
`correspond to those provided by JM1 of the JVT codec (shown in Table 3), with the exception that
`
`the SKIP mode is redefined to allow the representation of global and regional motion.” Id. at 17:43-
`
`47. Table 3 (shown below) depicts a SKIP mode followed by non-skip coding modes.
`
`b.
`
`Residual information
`
`
`
`The patent further describes that a decoder “uses the coding mode indication and motion
`
`vector information (if any) to form a prediction for the macroblock in question.” ’808 Patent at
`
`21:12-14. This prediction is done by forming an array of pixels by “copying corresponding pixel
`
`values from a region (or regions) of pixels in a reference frame.” Id. at 21:14-18. Then, the
`
`resulting prediction is “combined with the decoded prediction error information (if any)” to obtain
`
`an array of reconstructed pixel values. Id. at 21:18-24. These reconstructed pixel values are passed
`
`to video output and stored in “frame store” for future reference purposes. Id. at 21:25-31.
`
`Respondents selectively highlight discussion of Figure 6 (encoder) and Figure 7 (decoder)
`
`to allege that the ’808 Patent contemplates including residual information as part of its redefined
`
`skip coding mode. Resp. Br. at 20. However, Respondents ignore that none of their cited passages
`
`actually discuss the skip coding mode. See Resp. Br. at 20 (citing ’808 Patent at 16:51-55, 21:18-
`
`22), 14 (citing 16:22-31). Instead, the ’808 Patent provides that “[i]f a macroblock is encoded in
`
`10
`
`INV. NO. 337-TA-1379
`COMPLAINANTS’ REPLY CLAIM CONSTRUCTION BRIEF
`
`

`

`skip mode, no prediction error or motion vector information relating to the macroblock is included
`
`in the bit-stream.” ’808 Patent at 20:46-55. The ’808 Patent reinforces this point repeatedly. See,
`
`e.g., id. at 14:62-64 (“according to the invention,” no additional information in bitstream for skip
`
`coding mode must be sent to model global or regional motion); 15:7-10 (same); 15:66-67 (same);
`
`19:17-20 (same, albeit for an embodiment).
`
`This leads us back to Respondents’ claim differentiation theory. Given that the ’808 Patent
`
`states that residual information is not provided in skip coding mode, are dependent claims 30, 41,
`
`49, and 60, which recite “wherein no residual information is provided for the first segment in the
`
`encoded bitstream” rendered meaningless?
`
`One answer (discussed in the argument section below), is that this does not change the
`
`construction of “skip coding mode” because claim differentiation provides no basis to disregard
`
`clear lexicography or the unequivocal teachings of the specification. The inventor chose not to
`
`include “residual information” in their definition of skip coding code, and that is dispositive.
`
`Another answer is that Respondents confuse skip coding mode (the term up for
`
`construction) with the entire encoded bitstream (the subject of the negative limitation in these
`
`dependent claims). Recall that the skip coding mode as described in the ’808 Patent occurs and is
`
`signaled on a macroblock-by-macroblock basis within a larger bitstream. However, according to
`
`Respondents—albeit outside the scope of the ’808 Patent’s skip coding mode inventions—various
`
`residual information is possible, such as “residual information used with global motion and object
`
`motion” in an encoded bitstream. Resp. Br. at 8.
`
`Nokia need not identify or explain all conceivable forms of residual information—after all,
`
`residual information is simply not part of Nokia’s skip coding mode inventions and not part of the
`
`inventor’s definition of “skip coding mode.” Rather, it suffices to say that Respondents admit
`
`11
`
`INV. NO. 337-TA-1379
`COMPLAINANTS’ REPLY CLAIM CONSTRUCTION BRIEF
`
`

`

`residual information could exist in a bitstream untethered from a block-by-block coding mode
`
`process. As a technical matter, this is fatal to Respondents’ claim differentiation theory. The
`
`independent claims of the ’808 Patent (claims 1, 7, 10, and16) can all be satisfied with the redefined
`
`skip coding mode that does not include residual information for any skip mode macroblocks. At
`
`the same time, one would need further investigation to confirm whether “no residual information
`
`is provided for the first segment in the encoded bitstream” such that dependent claims 30, 41, 49,
`
`and 60 are satisfied. Namely, one would need to investigate the entire encoded bitstream rather
`
`than the skip coding mode indicated for a given macroblock. Therefore, given the difference
`
`between the independent and dependent claims, claims 30, 41, 49, and 60 cannot be “meaningless”
`
`or non-compliant with 35 U.S.C. ¶ 112 as Respondents contend.
`
`3. Whether Motion Information is Transmitted in or Derived From the
`Redefined Skip Coding Mode
`
`Respondents describe the ’808 Patent as “tak[ing] the ‘skip mode’ term from JM1 and then
`
`add[ing] its motion information technique.” Resp. Br. at 11 (citing ’808 Patent at 17:45-52, 19:28-
`
`51). Respondents later assert that the “the parties and Staff agree motion information can be
`
`6
`provided in skip coding mode.” Resp. Br. at 17. Both statements are inaccurate.5F
`
`First, the inventor was not merely taking just the words “skip mode” from JM1 and nothing
`
`else. The inventor used the skip coding mode of JM1 as a launch pad for lexicography. This makes
`
`sense because a POSITA familiar with JM1 would understand the context of the inventor’s
`
`redefinition of “skip coding mode.” Orchard Dec. at ¶¶ 20, 22. This is clear from the inventor’s
`
`definition at column 17 lines 43-53 that explains the modifications being made to the JM1 skip
`
`
`6 “Motion information” was not discussed on the claim construction meet and confer. Had
`Respondents asked Nokia whether it agreed “motion information can be provided in skip coding
`mode,” Nokia would have disagreed and directed Respondents to at least the ’808 Patent’s
`disclosures at 14:62-64; 15:7-10; and 15:66-67; 19:17-20; 19:46-61, and 20:6-25.
`12
`
`INV. NO. 337-TA-1379
`COMPLAINANTS’ REPLY CLAIM CONSTRUCTION BRIEF
`
`

`

`mode and that the redefined skip coding mode would be a more efficient substitute for JM1’s skip
`
`mode. ’808 Patent at 17:43-53 (lexicography), 10:23-39 (describing the recommendation under
`
`development); Orchard Dec. at ¶ 22.
`
`Second, Respondents’ statement that “motion information can be provided in skip coding
`
`mode” contradicts their admission on the very first page of their brief that “each of the ’808
`
`patent’s claims … describes a ‘skip coding mode’ wherein motion vectors are derived at the
`
`decoder rather than sent by the encoder.” Compare Resp. Br. at 1 (emphasis added) with id. at
`
`7 Because “motion vectors are derived at the
`17. The second statement is correct; the first is not.6F
`
`decoder rather than sent by the encoder,” as Respondents admit, including motion information
`
`would not make sense for the redefined skip coding mode. Moreover, “motion information” simply
`
`is not part of the inventor’s definition of skip coding mode.
`
`The notion that motion information is provided in skip coding mode confuses the signaling
`
`of skip coding mode in a bitstream with the subsequent motion vector generation and motion
`
`compensated prediction done by a decoder based on skip coding mode. Motion information is not
`
`“provided in skip coding mode” as Respondents allege on page 17 of their brief. Instead, as the
`
`claims and specification describe, after an encoder determines through motion estimation (e.g.,
`
`’808 Patent at 18:12-17) that it will use the redefined skip coding mode for a particular macroblock,
`
`the encoder includes an “indication” of the skip coding mode in the bitstream, which can then be
`
`transmitted from the encoder to a decoder. ’808 Patent at 18:17-23; see, e.g., Claim 1[D]
`
`(“providing…an indication of the skip coding mode…”). A decoder receives an indication of skip
`
`
`7 Based on Respondents’ description of the ’808 Patent on pages 11 and 17 of their opening brief,
`Nokia presumes Respondents are equating “motion information” with motion vectors describing
`the motion of macroblocks. Such motion vectors are not included in the bitstream as part of skip
`coding mode. See, e.g., 14:62-64; 15:7-10; and 15:66-67; 19:17-20; 19:46-61, and 20:6-25.
`13
`
`INV. NO. 337-TA-1379
`COMPLAINANTS’ REPLY CLAIM CONSTRUCTION BRIEF
`
`

`

`coding mode (see, e.g., Claim 7[A]) as part of a video bitstream a