NAUGHTY DOG, INC.v.MCRO, INC.

Patent Trial and Appeal Board

May 28, 2014

08942987 (P.T.A.B. May. 28, 2014)

Trials@uspto.gov Paper 9
Tel: 571-272-7822 Entered: May 28, 2014

UNITED STATES PATENT AND TRADEMARK OFFICE
_______________
BEFORE THE PATENT TRIAL AND APPEAL BOARD
_______________
NAUGHTY DOG, INC.
Petitioner
v.
MCRO, INC.
Patent Owner
_______________

Case IPR2014-00198
Patent 6,307,576 B1
_______________

Before BRIAN J. McNAMARA, RICHARD E. RICE, and
JEREMY M. PLENZLER, Administrative Patent Judges.

PLENZLER, Administrative Patent Judge.

DECISION
Denying Institution of Inter Partes Review
37 C.F.R. § 42.108

IPR2014-00198
Patent 6,307,576 B1

2

I. INTRODUCTION
A. Background
Naughty Dog, Inc. (“Petitioner”) filed a petition to institute an inter partes
review of claims 1-26 of U.S. Patent No. 6,307,576 B1 (“the ’576 patent”).
Paper 4 (“Petition” or “Pet.”). McRO, Inc. (“Patent Owner”) filed a preliminary
response. Paper 8 (“Prelim. Resp.”). The standard for instituting an inter partes
review is set forth in 35 U.S.C. § 314(a), which provides as follows:
THRESHOLD.—The Director may not authorize an inter partes
review to be instituted unless the Director determines that the
information presented in the petition filed under section 311 and any
response filed under section 313 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.
Petitioner contends that the challenged claims are unpatentable under
35 U.S.C. §§ 102 and 103 on the following specific grounds (Pet. 4-54):
References Basis Claims challenged
Peng
1
§ 102 1-26
Waters
2
§ 102 1-3, 7-16, 20-26
Kaneko
3
§ 102 1-3, 5, 7, 8, 10, 13-16, 18, 20, 21, 23, 26

1
Antai Peng, Speech Expression Modeling and Synthesis (Ph.D. dissertation,
Georgia Institute of Technology) (May 1996) (Ex. 1004) (“Peng”).
2
Keith Waters & Thomas M. Levergood, DECface: An Automatic Lip-
Synchronization Algorithm for Synthetic, CAMBRIDGE RES. LAB. TECH REPORT
SERIES (Sept. 23, 1993) (Ex. 1005) (“Waters”).
3
UK Patent Application No. GB 2231246 A, published Nov. 7, 1990 (Ex. 1006)
(“Kaneko”).
IPR2014-00198
Patent 6,307,576 B1

3

References Basis Claims challenged
Parke
4
§ 102 1-26
Peng and Waters § 103 1-26
Peng and Kaneko § 103 1-26
Peng and Parke § 103 1-26
Waters and Kaneko § 103 1-3, 5, 7-16, 18, 20-26
Waters and Parke § 103 1-26
Kaneko and Parke § 103 1-26
For the reasons given below, we deny institution on all grounds.
B. Additional Proceedings
Petitioner indicates that the ’576 patent is the subject of the following co-
pending federal district court case filed December 4, 2012: McRO, Inc., d.b.a.
Planet Blue v. Naughty Dog, Inc., Case No. 2:12-cv-10335-GW-FFM, pending in
the United States District Court for the Central District of California. Pet. 1-2.
Petitioner also indicates that the ’576 patent is asserted against Sony
Computer Entertainment America, LLC (of which Naughty Dog is a wholly owned
subsidiary) in a suit filed November 21, 2012, McRO, Inc., d.b.a. Planet Blue v.
Sony Computer Entertainment America, LLC and SCE Santa Monica Studio, Inc.,
Case No. 1:12-cv-01514-LPS, pending in the United States District Court for the
District of Delaware. Petitioner additionally indicates that the ’576 patent is
asserted against Sucker Punch Productions, LLC (also a wholly owned subsidiary
of Sony Computer Entertainment America, LLC) in a suit filed November 21,

4
Frederick I. Parke & Keith Waters, COMPUTER FACIAL ANIMATION (1996) (Ex.
1007) (“Parke”).
IPR2014-00198
Patent 6,307,576 B1

4

2012, McRO, Inc., d.b.a. Planet Blue v. Sucker Punch Productions, LLC, Case No.
1:12-cv-01515-LPS, pending in the United States District Court for the District of
Delaware. Id. at 2.
In addition to this Petition, Petitioner has filed a petition (IPR2014-00197)
challenging the patentability of Patent Owner’s U.S. Patent No. 6,611,278 B2 (“the
’278 patent”). Id. This decision will refer to the decision for IPR2014-00197 as
“the ’197 Decision.”
C. The Claimed Invention
The technology of the ’576 patent is identical to that described for the
related ’278 patent in the ’197 Decision.
5
The ’576 patent is titled “Method for
Automatically Animating Lip Synchronization and Facial Expression of Animated
Characters” and describes a set of rules that determine a stream of morph weight
sets when a sequence of timed phonemes is encountered. Ex. 1002, 3:63-66.
The ’576 patent explains that a “phoneme” is defined as “the smallest unit of
speech, and corresponds to a single sound.” Id. at 1:34-36. A “morph weight set”
is defined in the ’576 patent as “a set of values, one for each delta set, that, when
applied as described, transform the neutral model to some desired state, such as
speaking the ‘oo’ sound or the ‘th’ sound.” Id. at 4:38-41. The values (or morph
weights) usually range from 0 to 1, and are assigned to a delta set to modify the
geometry of the neutral model (i.e., a neutral mouth position). Id. at 1:63-66.
Delta sets are described in the ’576 patent as the deltas of each vertex on
each morph target relative to the neutral mouth position and are computed as a
vector from each vertex (n) on the neutral mouth position reference to each vertex
(n) on a morph target. Id. at 1:58-60. A morph target is described as a mouth

5
The ’278 patent is a continuation of the ’576 patent.
IPR2014-00198
Patent 6,307,576 B1

5

position displaced relative to the neutral mouth position and corresponding to a
phoneme or set of phonemes. Id. at 1:46-50. The ’576 patent explains that there is
one delta set for each morph target. Id. at 1:60-61.
An example sequence of timed phonemes (“the example phoneme
sequence”) is illustrated in a table at column 8, lines 42-51 of the ’576 patent and a
corresponding example morph weight set stream (“the example morph weight set
stream”) is illustrated in a table at column 8, line 56 through column 9, line 9. The
tables summarizing the example phoneme sequence and the example morph weight
set stream are reproduced below.

Example Phoneme Sequence Table
The example phoneme sequence table includes two columns, one for time and the
other for the phonemes corresponding to the times. The example phoneme
sequence is a simplified example, including only the delta sets relevant to the word
“hello.”

IPR2014-00198
Patent 6,307,576 B1

6

Example Morph Weight Set Stream Table
The example morph weight set stream table includes a column for time and a
column for each delta set. A series of morph weight sets are included in the
columns for the delta sets and form the morph weight set stream.
The ’576 patent describes rules that determine morph weight sets based on
encountered phonemes. Id. at 7:58-8:8. For example, if the phoneme “eh” is
encountered, the morph weight set (0,1,0,0,0,0) is used. Id. at 7:64-66. The ’576
patent also provides examples of rules that determine the morph weight set stream
as a function of phoneme sequence and times associated with the phoneme
sequence. Id. at 8:9-31. For example, transition rules are applied to determine
transition start and end times between phonemes as a function of the outgoing
phoneme’s end time and duration and the incoming phoneme’s start time and
duration. Id. at 8:9-14.
IPR2014-00198
Patent 6,307,576 B1

7

The application of the rules to an incoming phoneme sequence, such as the
example phoneme sequence, results in an output morph weight set stream, such as
the example morph weight set stream, indicating morph targets at times
corresponding to the phonemes of the incoming phoneme sequence. Id. at 7:59-
9:9. The ’576 patent explains that morph weight sets may be evaluated at times
between the start and end times in the morph weight set stream by interpolation.
Id. at 7:13-15.
Claim 1 illustrates the claimed subject matter and is reproduced below:
1. A method for automatically animating lip synchronization
and facial expression of three-dimensional characters
comprising:
obtaining a first set of rules that define output morph weight
set stream as a function of phoneme sequence and time of
said phoneme sequence;
obtaining a timed data file of phonemes having a plurality of
sub-sequences;
generating an intermediate stream of output morph weight
sets and a plurality of transition parameters between two
adjacent morph weight sets by evaluating said plurality
of sub-sequences against said first set of rules;
generating a final stream of output morph weight sets at a
desired frame rate from said intermediate stream of
output morph weight sets and said plurality of transition
parameters; and
applying said final stream of output morph weight sets to a
sequence of animated characters to produce lip syn-
chronization and facial expression control of said
animated characters.
IPR2014-00198
Patent 6,307,576 B1

8

D. Claim Construction
Consistent with the statute and the legislative history of the Leahy-Smith
America Invents Act,
6
the Board will interpret claims of an unexpired patent using
the broadest reasonable construction in light of the specification of the patent. See
Office Patent Trial Practice Guide, 77 Fed. Reg. 48,756, 48,766 (Aug. 14, 2012);
37 C.F.R. § 42.100(b). Claim terms are given their ordinary and customary
meaning, as would be understood by one of ordinary skill in the art in the context
of the disclosure. In re Translogic Tech., Inc., 504 F.3d 1249, 1257 (Fed. Cir.
2007). An inventor may rebut that presumption by providing a definition of the
term in the specification with reasonable clarity, deliberateness, and precision. In
re Paulsen, 30 F.3d 1475, 1480 (Fed. Cir. 1994).
1. “phoneme sequence”
The parties present the same opposing views on the construction of
“phoneme sequence” as those discussed in the ’197 Decision. Petitioner and
Patent Owner agree that a phoneme is the smallest unit of speech and corresponds
to a single sound. Pet. 11 (quoting Ex. 1002, 1:34-36); Prelim. Resp. 9-10 (citing
Ex. 1002, 1:35-37). Petitioner contends that a “phoneme sequence,” therefore, is a
sequence of phonemes or units of speech. Pet. 11. Patent Owner argues that
Petitioner’s proposed construction (a sequence of phonemes or units of speech) is
overly broad because the use of “or” reads “phoneme” out of the phrase. Prelim.
Resp. 9-10. Similar to the reasons discussed in the ’197 Decision, and for purposes
of this decision, we determine that a “phoneme sequence” is a sequence of the
smallest units of speech, i.e., a sequence of phonemes, which is consistent with the
portions of the specification cited by Petitioner and Patent Owner.

6
Pub. L. No. 112-29, 125 Stat. 284 (2011).
IPR2014-00198
Patent 6,307,576 B1

9

2. “morph weight set”
The parties agree on the definition of “morph weight set” as set forth in the
’576 patent, which is the same as the definition provided in the ’278 patent and that
discussed in the ’197 Decision. Pet. 11 (quoting Ex. 1002, 4:38-41); Prelim. Resp.
10 (quoting Ex. 1002, 4:38-42). Therefore, for purposes of this decision, we
determine that a “morph weight set” is a set of values, one for each delta set, that,
when applied as described, transform the neutral model to some desired state, as
defined by the ’576 patent. See Ex. 1002, 4:38-40.
3. “first set of rules”
Petitioner contends that the term “first set of rules” should be construed to
mean a set of conditional statements or correspondence between an input and an
output. Pet. 12. While not proposing a specific construction for this term, Patent
Owner argues that the surrounding claim language expressly requires that the “first
set of rules” “define output morph weight set stream as a function of a phoneme
sequence and time of said phoneme sequence.” Prelim. Resp. 12. As we discuss
further below, in view of the claim language, the specific definition of “first set of
rules” is not material to our decision in this proceeding. Therefore, we do not
construe this term separately.

II. ANALYSIS
A. Overview
Petitioner contends that claims 1-26 of the ’576 patent are unpatentable over
the prior art cited in the table above. Claims 1 and 14 are independent, with claims
2-13 depending from claim 1 and claims 15-26 depending from claim 14. In
support of the proposed challenges to these claims, Petitioner provides testimony
from Joseph Marks, Ph.D. (Ex. 1001; the “Marks Declaration”), which is similar to
IPR2014-00198
Patent 6,307,576 B1

10

the testimony provided in IPR2014-00197.
B. Proposed Ground of Anticipation over Peng (Ground 1)
1. Overview of Peng
Peng is a Ph.D. dissertation that addresses mapping phonemes to synthetic
facial images. Ex. 1004, x. The facial expression modeling in Peng is based on
the Facial Action Coding System (FACS). Id. at 14. FACS was developed prior to
Peng with the purpose of finding a mapping from speech phonemes to realistic
facial expressions. Id. FACS defines 46 action units (AUs) that describe facial
expressions. Id. at 15. Peng utilizes 11 AUs focusing on lower face actions for
facial expression modeling, which are summarized in Table 3-1, reproduced below.
Id. at 15-16.

Table 3-1 from Peng summarizes the AUs associated with facial expression
modeling, identifies the description of the action group (vertical, horizontal,
oblique, orbital, misc.) of the AUs, and provides a description of the facial feature
associated with each AU (e.g., raise upper lip). Peng discusses characterizing
IPR2014-00198
Patent 6,307,576 B1

11

these AUs as vectors and applying expression weight vectors to the AU vectors to
construct a desired facial expression. Id. at 16-18.
Peng discusses a key-frame animation technique where a text string is
converted to a stream of phonemes and their corresponding synthetic facial image
frames (key-frames). Id. at 52. These key-frames are determined based on the
AUs and weight vectors. Id. at 48. Peng explains that interpolation can be used to
smooth the transition between key-frames. Id. at 52-54. Peng additionally
discusses a phoneme parsing scheme to address differences in phoneme
appearance based on where the phoneme appears in a word or sentence. Id. at 55-
56.
2. Analysis
Petitioner contends that Peng discloses the limitations of claims 1-26. Pet.
14, 22-53. Similar to claim 1 of the ’278 patent, claim 1 of the ’576 patent recites
“obtaining a first set of rules that define output morph weight set stream as a
function of phoneme sequence and time of said phoneme sequence” (“limitation
1(a)”). Similar to claim 1, claim 14 of the ’576 patent recites “said first set of rules
defining output morph weight set stream as a function of phoneme sequence and
time of said phoneme sequence” and Petitioner’s contentions regarding this
limitation in claim 14 are the same as those presented regarding claim 1. See Pet.
14, 22, 38; Ex. 1001, 14-17, 61-62. For simplicity, the challenge will be discussed
with respect to claim 1.
Petitioner presents the same contentions regarding limitation 1(a) as those
discussed in the ’197 Decision. As noted above, Petitioner contends that the
claimed first set of rules requires “a set of conditional statements or
correspondence between an input and an output.” Pet. 12. Regardless of whether
the first set of rules requires a set of conditional statements or correspondence
IPR2014-00198
Patent 6,307,576 B1

12

between an input and an output, as Petitioner contends, the claim requires that the
rules define a morph weight set stream as a function of phoneme sequence and
times associated with said phoneme sequence.
Patent Owner acknowledges that Peng describes a rule base, but argues that
“there is no rule in Peng’s rule-base that defines a morph weight set stream as a
function of time of the phoneme sequence” as required by the claims. Prelim.
Resp. 18 (emphasis omitted). Petitioner contends that “in Peng, an output morph
weight set stream is generated by evaluating a plurality of phoneme sequences
against a set of rules,” and “[t]he rule base which is implemented in Peng’s finite
state machine is a function which produces the corresponding weight vector set
after evaluating the phonemic input against predetermined criteria.” Pet. 14.
Petitioner’s discussion of Peng relative to the ’576 patent does not identify any
disclosure in Peng that corresponds to a morph weight set stream being defined as
a function of time of said phoneme sequence. Petitioner’s claim chart also fails to
identify clearly what in Peng corresponds to the claimed “function of . . . time of
said phoneme sequence” in limitation 1(a). From Petitioner’s claim chart, it
appears that Petitioner may consider Peng’s weight vectors as corresponding to the
claimed morph weight set stream.
7
See Pet. 24 (citing Ex. 1004, 48, 50-52, 56-58).
However, as Patent Owner explains, Peng’s discussion of weight vectors on pages
48 and 50-51 does not indicate that the weight vectors are a function of phoneme
times or that they contain any time values. See Prelim. Resp. 18; Ex. 1004, 48, 50-
51.

7
This is an assumption, as Petitioner does not address clearly what discloses the
claimed morph weight set stream in Peng, and page 24 of the Petition is actually a
portion of the claim chart for a further claim limitation (“generating an output
morph weight set stream”).
IPR2014-00198
Patent 6,307,576 B1

13

Petitioner’s claim chart also cites to pages 14-17 of the Marks Declaration
for limitation 1(a). Pet. 22. However, after review of the cited portions of the
Marks Declaration, we see no further discussion of what in Peng allegedly
discloses the claimed “function of . . . time of said phoneme sequence” in
limitation 1(a). Rather, the cited portions of the Marks Declaration simply
reproduce the text from the portions of Peng cited in the Petition’s claim chart. See
Ex. 1001, 14-17.
In the challenge to claims 1 and 14, Petitioner has failed to specify what in
Peng corresponds to the claimed “function of . . . time of said phoneme sequence.”
Based on the Petition and evidence provided, we are not persuaded that Peng
discloses a first set of rules that defines a morph weight set stream as a function of
time of said phoneme sequence.
For these reasons, Petitioner has failed to establish a reasonable likelihood of
success on this challenge to claims 1-26.
C. Proposed Ground of Anticipation over Waters (Ground 2)
1. Overview of Waters
Waters is a report that addresses automatic synchronization of computer
generated faces with synthetic speech. Ex. 1005, Abstract. Figure 1 of Waters
illustrates a synchronization model, and is reproduced below.
IPR2014-00198
Patent 6,307,576 B1

14

Figure 1 of Waters is a schematic illustration of a system for converting input text
into synthesized audio and corresponding graphics of mouth deformations on a
display. The schematic illustration in Figure 1 includes input text, a text to speech
converter, a face generator, a mouth deformation calculator, and a viseme table.
Waters explains that mouth deformations corresponding to audio are based
on calculated displacement of mouth nodes at a series of times. Id. at 7.
Intermediate nodes can be interpolated at times between the calculated node
displacements. Id. Waters discusses determining node displacement based on
Newtonian physics. Id. at 8. Waters explains that this system provides for creation
of viseme tables that are based on the calculated mouth node displacements and
correlated to specific speech synthesizers. Id. at 18.
2. Analysis
Petitioner contends that Waters discloses the limitations of claims 1-3, 7-16,
and 20-26. Pet. 16, 22-53. As noted above, and similar to claim 1 of the ’278
IPR2014-00198
Patent 6,307,576 B1

15

patent, limitation 1(a) of claim 1 of the ’576 patent includes rules defining an
“output morph weight set stream” (i.e., a series of output morph weight sets).
Claim 14 of the ’576 patent includes the same limitation and Petitioner’s
contentions regarding this limitation in claim 14 are the same as those presented
regarding claim 1. See Pet. 16, 22, 38 and Ex. 1001, 69-72, 97. For simplicity, the
challenge will be discussed with respect to claim 1.
Petitioner presents the same contentions regarding limitation 1(a) as those
discussed in the ’197 Decision. Petitioner contends that “Waters relies on a rule
base of physical characteristics of mouth shapes (‘visemes’) corresponding to a
given phoneme input text” and that the “rules are implemented as an equation
(function) which calculates displacement of nodes on a parameterized facial
model.” Pet. 16. With respect to limitation 1(a), Petitioner’s claim chart cites to
the discussion of mouth deformation calculation and interpolation in Waters. Id. at
22-23 (citing Ex. 1005, 7-8). Petitioner’s claim chart also cites to pages 3, 5, 6,
and 11 of Waters and pages 69-72 of the Marks Declaration. See id. However, the
cited portions of the Marks Declaration simply reproduce the text from Waters
cited in the Petition’s claim chart. See Ex. 1001, 69-72.
Patent Owner argues that Petitioner fails to identify what in Waters
corresponds to the claimed output morph weight set streams. Prelim. Resp. 24.
Patent Owner further argues that regardless of what Petitioner may consider as the
claimed output morph weight set, nothing in Waters discloses this feature. Id. at
24, 27. It is unclear from the Petition whether Petitioner considers the calculated
node displacement (xi(t)), the intermediate interpolation position (x(s)), or some
other feature on pages 3, 5-8, and 11 of Waters to be an output morph weight set.
It is not apparent why the cited portions of Waters disclose an output morph weight
IPR2014-00198
Patent 6,307,576 B1

16

set. Based on the Petition and the evidence provided, we are not persuaded that
Waters discloses an output morph weight set.
For these reasons, Petitioner has failed to establish a reasonable likelihood of
success on this challenge to claims 1-3, 7-16, and 20-26.
D. Proposed Ground of Anticipation over Kaneko (Ground 3)
1. Overview of Kaneko
Kaneko is a foreign Patent Application directed to synthesizing an image of
a face to represent changes in mouth shape based on speech output. Ex. 1006, 1:2-
7. Figure 4 illustrates an embodiment of the image synthesis system of Kaneko
and is reproduced below.

Figure 4 of Kaneko is a block diagram illustrating a system for converting input
text to an output picture corresponding to the input text. Figure 4 includes speech
synthesizer 1 that receives input text. Time adjuster 2 controls an input to picture
IPR2014-00198
Patent 6,307,576 B1

17

generator 6 based on the duration of the phoneme provided by speech synthesizer
1. Id. at 9:7-11.
Converter 3 converts the phoneme from speech synthesizer 1 to a mouth-
shape feature corresponding to the phoneme. Id. at 9:17-19. Examples of mouth-
shape features include degree of mouth opening, degree of roundness of lips,
height of lower jaw, and degree to which the tongue is seen. Id. at 9:19-24.
Correspondence between the vocal sound feature and the mouth-shape feature is
determined based on observations of how a person actually utters each phoneme.
Id. at 9:24-26. Conversion table 4 converts the mouth-shape feature to a mouth-
shape parameter value representing a concrete mouth shape (i.e., a specific
displacement of points P1-P8 defining mouth position). Id. at 10:13-11:3. The
mouth-shape parameter values are predetermined based on mouth-shape
measurements of a person actually uttering each phoneme. Id. at 11:3-5.
Transition detector 8 detects when a transition between phonemes occurs. Id. at
14:23-15:11. Mouth-shape parameter modifier 7 obtains intermediate parameter
values between consecutive mouth shape parameters for intermediate frames. Id.
at 15:14-20.
2. Analysis
Petitioner contends that Kaneko discloses each of the limitations in claims 1-
3, 5, 7, 8, 10, 13-16, 18, 20, 21, 23, and 26. Pet. 18-19, 22-53. As previously
indicated, and similar to the ’278 patent, limitation 1(a) from claim 1 of the ’576
patent includes rules defining an “output morph weight set stream.” Claim 14 of
the ’576 patent includes the same limitation and Petitioner’s contentions regarding
this limitation in claim 14 are the same as those presented regarding claim 1. See
Pet. 18-19, 23, 38-39; Ex. 1001, 102-103, 118. For simplicity, the challenge will
be discussed with respect to claim 1.
IPR2014-00198
Patent 6,307,576 B1

18

Petitioner presents the same contentions regarding limitation 1(a) as those
discussed in the ’197 Decision. Petitioner contends that “Kaneko uses a
conversion table which represents a rule base specifying mouth shapes
corresponding to phonemes.” Pet. 18. With respect to limitation 1(a), Petitioner’s
claim chart cites to Kaneko’s discussion of the operation of speech synthesizer 1,
which postulates the use of an existing speech rule synthesizing method that
employs a Klatt type format speech synthesizer for matching mouth shape
generation. Id. at 23 (citing Ex. 1006, 8:6-9:2). Kaneko provides no detailed
description of the speech synthesizer. Petitioner’s claim chart also cites to pages
102-103 of the Marks Declaration. However, again, the cited portions of the
Marks Declaration simply reproduce the text from Kaneko cited in the Petition’s
claim chart. See Ex. 1001, 102-103.
Patent Owner argues that “there is no set of rules in Kaneko’s synthesizer 1
that defines a morph weight set stream as required by claim 1.” Prelim. Resp. 33.
Indeed, page 8, line 6 through page 9, line 2 of Kaneko generally explains that
“speech synthesizer 1 synthesizes a speech output corresponding to an input
sentence” and discusses various speech synthesizing methods. There is no
discussion of rules or any disclosure that appears to correspond to an output morph
weight set in this portion of Kaneko.
The additional portions of the claim chart related to claim 1 also fail to
identify what allegedly corresponds to rules defining an output morph weight set
stream in Kaneko. See Pet. 22-26. To the extent Petitioner considers Kaneko’s
conversion table 4 as corresponding to the claimed rules, Petitioner does not allege,
and offers no explanation as to why, conversion table 4 defines a morph weight set
stream. The only portion of the claim chart for claim 1 that references the
conversion table, discussed earlier in the Petition (see Pet. 18) as corresponding to
IPR2014-00198
Patent 6,307,576 B1

19

the claimed rules, is the portion directed to limitation (c) in claim 1: “generating
an intermediate stream of output morph weight sets . . . by evaluating said plurality
of sub-sequences against said first set of rules.” Pet. 24 (citing Ex. 1006, 11:6-10).
However, the cited portion of Kaneko simply states that “[i]n response to the
mouth-shape feature corresponding to the phoneme concerned, provided from the
speech feature to mouth-shape feature converter 3, the unit 5 refers to the
conversion table 4 to read out therefrom a set of values of mouth-shape parameters
for the phoneme.” Ex. 1006, 11:6-10. We are not persuaded that this discloses
rules defining an output morph weight set stream.
For these reasons, Petitioner has failed to establish a reasonable likelihood of
success on this challenge to claims 1-3, 5, 7, 8, 10, 13-16, 18, 20, 21, 23, and 26.
E. Proposed Ground of Anticipation over Parke (Ground 4)
1. Overview of Parke
Parke is a book about computer facial models, computer generated facial
images, and facial animation, and is intended to summarize “the diverse
developments in facial animation.” Ex. 1007, xiii. Parke discusses advertising,
film, gaming, medicine, teleconferencing, and avatar applications. Id. at 5-7.
Parke discusses a variety of facial animation techniques. Id. at 105-146. Parke
also discusses a variety of methods for speech synchronized animation. Id. at 259-
284.
2. Analysis
Petitioner contends that Parke discloses each of the limitations in claims 1-
26. Pet. 20-53. As noted above, and similar to claim 1 of the ’278 patent,
limitation 1(a) from claim 1 of the ’576 patent includes rules defining an “output
morph weight set stream.” Claim 14 of the ’576 patent includes the same
limitation and Petitioner’s contentions regarding this limitation in claim 14 are the
IPR2014-00198
Patent 6,307,576 B1

20

same as those presented regarding claim 1. See Pet. 20-21, 23, 39; Ex. 1001, 123-
124, 157. For simplicity, the challenge will be discussed with respect to claim 1.
Petitioner presents the same contentions regarding limitation 1(a) as those
discussed in the ’197 Decision. Petitioner contends that “[c]oncepts discussed [in
Parke] include viseme table lookup, speech synthesizer output parameters, key
framing, and interpolation. For example, key frames (mouth positions)
corresponding to phoneme input may be obtained from a viseme table.” Pet. 20.
With respect to limitation 1(a), Petitioner’s claim chart refers to pages 270 and 273
of Parke. Pet. 23. The cited portions of Parke generally discuss generation of a
phoneme sequence from input text, determination of timing for facial animation
based on input text and corresponding phonemes, and synthesis by rule algorithms.
Ex. 1007, 270, 273.
Patent Owner argues that “Petitioner never identifies what in Parke allegedly
corresponds to the claimed output morph weight set stream, or what in Parke
allegedly corresponds to the claimed first set of rules.” Prelim. Resp. 43. Patent
Owner argues that although pages 270 and 273 of Parke discuss the use of
phoneme sequences, synthesis by rule algorithms, and timing, Parke does not
disclose the claimed output morph weight set stream. Id.
Based on the Petition and evidence provided, we are unable to determine
what in Parke Petitioner considers as corresponding to the claimed rules defining a
morph weight set stream. For example, Petitioner does not offer any explanation
as to why Parke’s discussion of phoneme sequences, synthesis by rule algorithms,
and timing discloses rules defining an output morph weight set stream. The cited
portions of the Marks Declaration simply reproduce the text from Parke cited in the
Petition’s claim chart. See Ex. 1001, 143-144. We are not persuaded that the
general discussion found on pages 270 and 273 of Parke discloses this limitation,
IPR2014-00198
Patent 6,307,576 B1

21

nor are we persuaded that Parke discloses rules defining an output morph weight
set stream based on Petitioner’s general summary of “concepts discussed” in Parke
(see Pet. 20).
For these reasons, Petitioner has failed to establish a reasonable likelihood of
success on this challenge to claims 1-26.
F. Proposed Obviousness Grounds 5-10
1. Overview
Petitioner proposes a number of obviousness challenges based on various
combinations of Peng, Waters, Kaneko, and Parke. Pet. 5, 21-58.
2. Analysis
Petitioner presents similar contentions regarding these challenges, at least
with respect to limitation 1(a) and the similar limitation in claim 14, to those
discussed in the ’197 Decision. Similar to the contentions presented regarding the
’278 patent in IPR2014-00197, Petitioner contends that the ’576 patent claims
“merely recite the combination of ‘prior art elements according to known methods
to yield predictable results’ and/or the ‘[u]se of known technique[s] to improve
similar devices (methods, or products) in the same way.’ MPEP § 2143 (A, C).”
Id. at 21.
Patent Owner argues that
(1) Petitioner NEVER explains what is allegedly lacking in the base
reference, (2) Petitioner NEVER points out the differences between
the claimed invention and the base reference or cited art, (3) Petitioner
NEVER explains how the base reference should be allegedly
modified, (4) Petitioner NEVER explains what modification would
have been obvious, and (5) Petitioner NEVER explains what
modifications to each base reference would have allegedly been made.
Prelim. Resp. 47-48. We agree.
IPR2014-00198
Patent 6,307,576 B1

22

Petitioner does not identify any specific features of the cited references to be
included in the proposed combinations and does not provide any further rationale
for the combinations. See id. at 22-53; Ex. 1001, 163-543. The Petition’s claim
chart simply includes citations to the claim chart in the Marks Declaration and to
additional rationale for a general combination of the references in the Marks
Declaration. Id. at 22-53. The claim charts in the Marks Declaration do not
identify specifically which elements are to be substituted in the various proposed
combinations or how the references are to be modified. See Ex. 1001, 164-543.
The claim charts only cite to the references in the combination proposed in the
challenge without specifying which reference is relied on to meet each limitation.
For example, in proposed Ground 5 (challenge based on the combination of
Peng and Waters), the Petition cites the same portions of Peng relied on in
proposed Ground 1 (anticipation challenge based on Peng) and the same portions
of Waters relied on in proposed Ground 2 (anticipation challenge based on Waters)
for limitation 1(a). Pet. 22. The Petition additionally cites to pages 164-165 of the
Marks Declaration. Id. However, the cited portion of the Marks Declaration
simply includes the text from the portion of Peng cited in the Petition’s claim chart
with the addition of “See also Waters at, e.g., pp. 3, 5, 6, 7, 11.” Ex. 1001, 165-
168.
The rationale for the proposed combinations also fails to identify specifically
any proposed modifications. Instead, for each obviousness challenge, the Marks
Declaration summarizes the references relied on, and then concludes that
To a skilled artisan, this combination would have yielded a
predictable result—namely, a more desirable system for not only
utilizing an input and other data to automatically into animate lip
synchronization, but also to modify a facial expression of a character
based on emotional data. The implementation of such a system would
be more efficient, resulting in an implicit, common sense motivation
IPR2014-00198
Patent 6,307,576 B1

23

to combine the references to create a computer animation system for
automatically animating lip synchronization and facial expression of a
character, such as a three-dimensional character.
Id. at 237-238, 312, 386, 443, 501, 543-544.
Petitioner has failed to resolve any differences between the claimed
invention and the cited references, identify any specific proposed modifications to
the references, or explain persuasively why one skilled in the art would have made
any specific modifications to the references relied on in the challenges in Grounds
5-10. Based on the Petition and evidence provided, we are not persuaded that any
of the obviousness challenges (Grounds 5-10) cures the deficiencies in the
challenges based on Peng, Waters, Kaneko, and Parke (Grounds 1-4) discussed
above. Thus, we are not persuaded that Petitioner has established a reasonable
likelihood of succeeding on any of these challenges to claims 1-26.
G. Petitioner’s Challenge to Dates of Peng and Waters
In addition to the arguments discussed above, Patent Owner additionally
argues that Petitioner has not established that Peng and Waters are prior art printed
publications. Prelim. Resp. 3-9. For example, Patent Owner argues that although
“Peng is an alleged ‘thesis’ . . . petitioner has submitted no evidence establishing
that Peng was sufficiently publicly accessible prior to the critical date (October 2,
1997) to qualify as a printed publication.” Id. at 4. Patent Owner provides similar
arguments regarding Waters, asserting that “there is no evidence that Waters was
ever disseminated outside of Digital Equipment Corporation (DEC), and there is
no evidence that anyone outside of DEC exercising reasonable diligence would
have located Waters prior to the critical date.” Id. at 8. Patent Owner argues that
neither Peng nor Waters should be considered as prior art printed publications for
these reasons. Id. at 9.
Because we have determined that Petitioner fails to meet its burden
IPR2014-00198
Patent 6,307,576 B1

24

regarding Peng and Waters meeting the limitations of the challenged claims, we
need not address whether Peng or Waters is a prior art printed publication.
Therefore, the issues raised in Patent Owner’s Preliminary Response regarding
whether Peng and Waters are prior art printed publications are moot.

III. SUMMARY
Petitioner has failed to demonstrate that there is a reasonable likelihood of
prevailing on any of the challenges to the patentability of claims 1-26 of the ʼ576
patent.
The Petition is denied as to all grounds proposed.

IV. ORDER
For the reasons given, it is
ORDERED that the Petition is denied.

IPR2014-00198
Patent 6,307,576 B1

25

For PETITIONER:

Tawni Wilhelm
twilhelm@shb.com

Jonathan Zerger
jzerger@shb.com

For PATENT OWNER:

Joseph Rhoa
jar@nixonvan.com

Updeep Gill
usg@nixonvan.com