`
`
`
`Filed on behalf of Securus Technologies, Inc.
`By:
`Justin B. Kimble (JKimble-IPR@bcpc-law.com)
`
`Jeffrey R. Bragalone (jbragalone@bcpc-law.com)
`
`Daniel F. Olejko (dolejko@bcpc-law.com)
`
`Nicholas C. Kliewer (nkliewer@bcpc-law.com)
`
`Bragalone Conroy P.C.
`
`2200 Ross Ave.
`
`Suite 4500 – West
`
`Dallas, TX 75201
`
`Tel: 214.785.6670
`
`Fax: 214.786.6680
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`GLOBAL TEL*LINK CORPORATION,
`Petitioner,
`
`v.
`
`SECURUS TECHNOLOGIES, INC.,
`Patent Owner.
`
`Case IPR2016-01220
`Patent 9,007,420 B1
`
`
`
`
`
`PATENT OWNER’S RESPONSE
`
`
`
`Mail Stop PATENT BOARD
`Patent Trial and Appeal Board
`U.S. Patent & Trademark Office
`P.O. Box 1450
`Alexandria, Virginia 22313-1450
`
`
`
`
`
`
`
`Case IPR2016-01220
`Patent 9,007,420
`Table of Contents
`
`I.
`
`INTRODUCTION .............................................................................................. 5
`
`A. Grounds in the Petition ............................................................................... 6
`B. Background Information and the Solution Provided by the ‘420 Patent. .. 6
`II. CLAIM CONSTRUCTION ............................................................................. 10
`
`A. An “actual face” as Used in Claims 1-20 Refers to the User’s Physical
`Face and Not a Facsimile of a Face Such as a Photograph. ..................... 11
`Claim Differentiation Provides a Presumption That Terms are Different .................. 12
`
`The ‘420 Patent Specification Provides Meaning and Context .................................. 13
`
`A Dictionary Definition Provides Insight into the Plain and Ordinary Meaning ....... 16
`
`Dr. Kakadiaris provides informed testimony .............................................................. 16
`
`Prosecution History Disclaimer of the Term’s Scope................................................. 17
`
`B. The Construction of “feature detection process” and “three dimensional
`feature detection process” Cannot Ignore and Must be Reasonable in
`Light of the Teachings of the Specification. ............................................ 20
`“performing a three-dimensional (3D) facial recognition process on the
`[first/second] image … to identify the user” should be construed as “a
`three-dimensional feature detection process followed by a process for
`comparing the facial landmarks of the face present in the image with the
`facial landmarks of a known likeness of the user and determining
`whether the facial landmarks of the face present in the image match the
`facial landmarks of the known likeness of the user.” ............................... 23
`Petitioner’s Construction of “Electronic Visitation Session” is
`Unreasonably Broad. ................................................................................ 25
`III. ARGUMENT ................................................................................................... 29
`
`D.
`
`C.
`
`2
`
`
`
`Case IPR2016-01220
`Patent 9,007,420
`A. The Combination of Torgersrud and Kenoyer Fails to Disclose the
`Limitations Requiring Determining or Verifying the Presence of an
`“Actual Face” in Independent Claims 1 and 11. ...................................... 29
`B. Zhang Does Not Disclose a “three-dimensional 3D feature detection” or
`“a three-dimensional (3D) facial recognition process.” ........................... 37
`C. Torgersrud Fails to Disclose Capturing an Image, with an Image Capture
`Device, of a User “in Response to” the “Request to Initiate an Electronic
`Visitation Session” in Independent Claims 1, 11, and 21. ....................... 40
`Persons of Ordinary Skill Would Not Have Combined Zhang With
`Kenoyer or Torgersrud. ............................................................................ 42
`Petitioner Fails to Consider Claim 21 as a Whole. .................................. 52
`E.
`IV. CONCLUSION ................................................................................................ 54
`
`D.
`
`3
`
`
`
`Case IPR2016-01220
`
`Patent 9,007,420
`
`PATENT OWNER’S EXHIBIT LIST
`
`
`Exhibit Description—
`
`
`
`2001
`
`Dictionary of Computer and Information Technology 272 (2013)
`
`2009
`
`Xerxes Mazda & Fraidoon Mazda, The Focal Illustrated Dictionary
`
`of Telecommunications 555 (1999)
`
`
`
`
`
`
`
`Case IPR2016-01220
`
`Patent 9,007,420
`
`I.
`
`INTRODUCTION
`
`Pursuant to 35 U.S.C. § 316(a)(8) and 37 C.F.R. § 42.120, Patent Owner
`
`Securus Technologies, Inc. (“Securus”) hereby responds to the Petition (Paper 2)
`
`(“Petition”)
`
`filed by Global Tel*Link Corporation (“GTL” or “Petitioner”)
`
`challenging claims 1-21 of U.S_ Patent No. 9,007,420 (EX. 1001) (the “‘420 Patent”)
`
`with respect to Grounds 1 and 2 which the Board instituted Inter Partes Review as
`
`summarized below:
`
`2-9, 12-19 2
`
`References Combined
`Torgersrud and Kenoyer
`Torgersrud, Kenoyer, and 21
`Zhan
`
`Claims
`
`Dependent
`
`Claims
`
`10, 20
`
`Petitioner bears “the burden of proving a proposition of unpatentability by a
`
`preponderance of the evidence-” 35 U.S.C. § 316(e). For the reasons set forth herein,
`
`Petitioner has failed to meet its burden to show by a preponderance of the evidence
`
`because (1) Petitioner’s grounds of unpatentability rely on incorrect claim
`
`constructions; (2) the prior art lacks material claim limitations; and (3) one having
`
`ordinary skill in the art would not have been motivated to combine references as
`
`Petitioner suggests; and (4) Petitioner uses impermissible hindsight reconstruction
`
`by failing to consider the differences between the claims and the prior art and failing
`
`to analyze the claims as a whole.
`
`
`
`Case IPR2016-01220
`Patent 9,007,420
`A. Grounds in the Petition
`
`The Petition relies on the combination of Torgersrud (U.S. Patent App. Pub.
`
`No. 2012/0262271 A1) and Kenoyer (U.S. Patent No. 8,218,829) for allegedly
`
`rendering obvious independent claims 1 and 11 of the ‘420 Patent under 35 U.S.C.
`
`§ 103 under Ground 1. Petitioner additionally relies upon adding Zhang (U.S. Patent
`
`No. 7,436,988) under Ground 2. Pet. at 12.
`
`As discussed in more detail below, none of the references above, either
`
`separately or in combination, disclose all limitations in the independent claims,
`
`including, e.g., the detection of the presence of an “actual face” or the capture of a
`
`user’s image in response to a request to initiate an electronic visitation session. Thus,
`
`the Petition does not demonstrate a reasonable likelihood that any of the proposed
`
`grounds of unpatentability will succeed for any claim of the ‘420 patent.
`
`B. Background Information and the Solution Provided by the
`‘420 Patent.
`
`A controlled-environment facility such as a prison may provide various
`
`options for inmates to communicate with visitors and other parties outside the prison.
`
`One of these options includes video visitation. These outside parties (e.g., family
`
`members) are sometimes located long distances from the inmate, making travel to
`
`the prison onerous or impractical. More frequently, outside parties may be allowed
`
`to visit with an inmate remotely using a personal computer that has an image capture
`
`device such as a webcam. This is sometimes referred to as “at home visitation,” and
`
`6
`
`
`
`Case IPR2016-01220
`Patent 9,007,420
`occurs via “electronic visitation sessions.” However, inmates are typically restricted
`
`to receiving visitation only from approved persons. ‘420 Patent at 1:14-34.
`
`When an individual visits an inmate in person, the individual’s identity may
`
`easily be determined by providing identification documents to staff of the controlled-
`
`environment facility for verification. Id. The identification documents may include
`
`a photo-ID such as a driver’s license or the like and the staff members may cross-
`
`reference the individual’s name with a list of individuals on the inmate’s approved
`
`visitor list.
`
`But identification of visitors is more difficult with telephone or video
`
`visitation. Id. at 1:27-35. This is especially true in the case of at-home visitation
`
`because there is no staff to physically verify the physical presence and identity of
`
`the authorized visitor. The individual may provide a personal identification number,
`
`phrase, or passcode, but it is often difficult to ascertain whether the person providing
`
`the identifying information is in fact the authorized visitor without visual
`
`confirmation of the person’s identity. Id. Unauthorized parties may attempt to defeat
`
`or circumvent verification of a user’s identity. Id. For example, an authorized visitor
`
`may pass identification information to unauthorized individuals so that they may
`
`pose as an authorized visitor for the electronic visitation. Id. Prior attempts to thwart
`
`this unauthorized activity included the use of facial recognition software to verify
`
`the identity of the user.
`
`7
`
`
`
`Case IPR2016-01220
`Patent 9,007,420
`Key to the significance of the ‘420 Patent, the inventor recognized that these
`
`typical facial recognition verification techniques are susceptible to circumvention.
`
`See ‘420 Patent at 1:27-35; 8:66-9:12. In particular, inmates or unauthorized outside
`
`parties could present a photograph of an authorized person in front of the visitation
`
`camera when prompted to verify the user’s identity. The ‘420 Patent particularly
`
`addresses and solves these problems, in part, by detecting and verifying the genuine
`
`presence of a user’s physical face (i.e., their “actual face”) in the image captured by
`
`the visitation camera. See, e.g., id., Abstract (providing a device “to verify that an
`
`actual face was present”). As such, the ‘420 Patent provides a way to confirm that
`
`an actual person, as opposed to a spoof facsimile, is taking part in a video visitation
`
`session.
`
`The specification of the ‘420 Patent describes the methods used in the claims
`
`that specifically require detection of an actual face (i.e., Claims 1 and 20). ‘420
`
`Patent at 8:66-9:12. In particular, the specification teaches that the feature detection
`
`process used to verify that “the person is an actual person and not simply a
`
`photograph presented to trick the system” may be a 3D feature detection process that
`
`can identify three-dimensional characteristics of an image, (Id., Fig. 4, 403)
`
`including measurements of features of a face at a plurality of points on the image
`
`and examining changes in the measurements from frame to frame of the captured
`
`video image. Id.; see also Claim 4. Alternatively, a “plurality of still frame
`
`8
`
`
`
`Case IPR2016-01220
`Patent 9,007,420
`photograms may be captured and differences in measurements may be calculated to
`
`determine if the presented person is an actual person or a photograph.” Id.; see also
`
`Claim 3.
`
`Other claims require further security measures such as facial recognition to
`
`affirmatively identify a user as being an authorized user. See id., Fig. 5, 504; 10:45-
`
`49. The facial recognition may include three-dimensional facial recognition. See id.
`
`at 9:34-36; 9:55-58; Claim 21.
`
`Each embodiment also includes certain triggers and event timing in order
`
`perform the user verification. Id. at 11:4-12:53. For example, each claimed
`
`embodiment includes receiving a request to initiate an electronic visitation session
`
`and capturing a user image in response to the request. Id. Furthermore, the claimed
`
`embodiments involve verification of the actual face at particular times during the
`
`process, including continued confirmation of the actual face based on a second
`
`image captured during the electronic visitation session. Id.
`
`The Petition attempts to characterize the ‘420 Patent as merely describing
`
`“using face recognition during a video conference.” Pet. at 9 (“Securus claims to
`
`have invented using face recognition during a video conference with an inmate.”).
`
`But, as discussed above and further analyzed below, the Petition grossly
`
`mischaracterizes the claimed embodiments of the ‘420 Patent. While “face
`
`recognition” is an element of some of the claimed embodiments, none of the prior
`
`9
`
`
`
`Case IPR2016-01220
`Patent 9,007,420
`art references disclose, teach, or suggest, separately or in combination, all of the
`
`limitations of any of the claims. Petitioner thus fails to prove prima facie case of
`
`obviousness and therefore fails to meet its burden to show invalidity of any of the
`
`claims of the ‘420 Patent.
`
`II. CLAIM CONSTRUCTION
`In the Decision to Institute (Paper 8), the Board preliminarily construed the
`
`terms “electronic visitation session” and “feature detection process.” Patent Owner
`
`respectfully suggests that the Board consider the arguments and highlighted
`
`evidence presented below and adopt Patent Owner’s revised constructions of these
`
`terms. Additionally, there appears to be contention between Petitioner and Patent
`
`Owner’s interpretation of the following claim terms: “actual face,” “3D feature
`
`detection process,” and “3D facial recognition.” Thus, the Board should construe
`
`these terms as well.
`
`Even under the broadest reasonable interpretation standard applicable here,
`
`claim constructions must be consistent with the patent specification. Servicenow,
`
`Inc. v. Hewlett-Packard Co., No. IPR2015-00702, Paper No. 12, Decision Denying
`
`Institution of Inter Partes Review at 6 (PTAB Aug. 21, 2015) (citing Microsoft Corp.
`
`v. Proxyconn, Inc., 789 F.3d 1291, 1298 (Fed. Cir. 2015)). As the Board has
`
`acknowledged, a claim’s construction “should always be read in light of the
`
`specification and teaching in the underlying patent,” and “cannot be divorced from
`
`10
`
`
`
`Case IPR2016-01220
`Patent 9,007,420
`the specification.” Id. Indeed, “the specification ‘is the single best guide to the
`
`meaning of a disputed term.’” In re Translogic Tech., at 1257 (Fed. Cir. 2007)
`
`(quoting Phillips v. AWH Corp., 415 F.3d 1303, 1315 (Fed. Cir. 2005)). Because the
`
`Petitioner’s constructions are unreasonable in light of the specification they cannot
`
`be correct. Servicenow, Inc., No. IPR2015-00702, Paper No. 12 at 6 (quoting
`
`Microsoft Corp., 789 F.3d at 1298) (“The Board . . . may not ‘construe claims during
`
`IPR so broadly that its constructions are unreasonable under general claim
`
`construction principles.’”).
`
`A. An “actual face” as Used in Claims 1-20 Refers to the User’s Physical
`Face and Not a Facsimile of a Face Such as a Photograph.
`
`Claims 1-20 use the term “actual face” whereas independent claim 21 uses the
`
`term “face.” The term “actual face” should be construed as referring to the user’s
`
`physical face and not a facsimile of a face such as a photograph because: (1) claim
`
`differentiation provides a presumption that the claim terms are different; (2) the
`
`specification of the ‘420 Patent supports the tendered construction and informs the
`
`meaning of the claim terms; (3) the dictionary definition of the term provides a plain
`
`and ordinary beaning meaning of the claim language; (4) Patent Owner’s expert,
`
`Dr. Kakadiaris provides informed testimony that one having ordinary skill in the art
`
`would have understood the broadest reasonable interpretation of the claim term to
`
`be the meaning provided above; and (5) the prosecution history (Ex.1002) provides
`
`11
`
`
`
`Case IPR2016-01220
`Patent 9,007,420
`several examples of Patent Owner’s prosecution history disclaimer of the term’s
`
`scope consistent with Patent Owner’s construction.
`
` Claim Differentiation Provides a Presumption That Terms are
`Different
`
`One of the classic canons of claim construction is that different claim terms
`
`in a patent are presumed to have different meanings. Comark Commc’ns, Inc. v.
`
`Harris Corp., 156 F.3d 1182, 1187 (Fed. Cir. 1998) (“There is presumed to be a
`
`difference in meaning and scope when different words or phrases are used in separate
`
`claims. To the extent that the absence of such difference in meaning and scope would
`
`make a claim superfluous, the doctrine of claim differentiation states the
`
`presumption that the difference between claims is significant.”) (citing Tandon
`
`Corp. v. United States Int’l Trade Comm’n, 831 F.2d 1017, 1023 (Fed. Cir. 1987)).
`
`Independent claims 1 and 11 both require validation of an actual face.
`
`Independent claim 21, however, uses the claim term “face,” thus creating a
`
`presumption that the term “face” has broader scope than the term “actual face.” Here,
`
`the term “actual” serves to distinguish the term “actual face” i.e. a corporeal face
`
`from merely a “face.” As explained below with reference to the specification and the
`
`prosecution history, the term “actual” is specifically provided to add the requirement
`
`that the claims detect whether an unauthorized user is spoofing the system by
`
`presenting a photograph of an authorized face instead of an “actual” (i.e., physical,
`
`genuine, authentic, corporeal) face as taught by the ‘420 patent.
`
`12
`
`
`
`Case IPR2016-01220
`Patent 9,007,420
`Generally, construction of a term is not allowed such broad play as to
`
`effectively eliminate that element in its entirety. See Warner-Jenkinson Co. v. Hilton
`
`Davis Chem. Co., 520 U.S. 17, 29 (1997). Recently, the Federal Circuit vacated and
`
`remanded a final written decision in an inter partes review for reading a claim term
`
`so broadly that it “reads that limitation out of the claim.” Los Angeles Biomedical
`
`Research Inst. at Harbor–UCLA Med. Ctr. v. Eli Lilly & Co., No. 2016-1518, ---
`
`F.3d ----, 2017 WL 765812, at *6 (Fed. Cir. Feb. 28, 2017) (holding that the plain
`
`and ordinary meanings of the words in the term were the broadest reasonable
`
`interpretation).
`
` The ‘420 Patent Specification Provides Meaning and Context
`Additionally, the specification of the ‘420 Patent makes a clear and consistent
`
`use of the term “actual face.” Besides appearing in the claims themselves, the claim
`
`term “actual face” appears in two paragraphs (‘420 Patent at 8:66-9:16) of the
`
`Detailed Description section, Figure 4, as well as in the Abstract and Summary
`
`sections of the specification. The terms appearance in the Abstract and in two of the
`
`three independent claims highlights the importance of the term to the ‘420 Patent.
`
`The relevant text of the ‘420 Patent reproduced below:
`
`The captured image may include either a still frame photograph or a
`video image. A single video frame may be captured or a plurality of
`video frames may be recorded. The image may be captured by
`
`13
`
`
`
`Case IPR2016-01220
`Patent 9,007,420
`camera 114, image capture device 395, or camera 205 in various
`embodiments.
`
`In an embodiment, a processor 310 of visitation system 130 may
`perform a three-dimensional (3D) feature detection process on the
`captured image to verify that an actual face is present in the image as
`shown at block 403. A 3D feature detection process may identify three-
`dimensional characteristics of an image, including measurements of
`features of a face at a plurality of points on the image. In certain
`embodiments, changes in the measurements from frame to frame of a
`video image may indicate that the person is an actual person and not
`simply a photograph presented to trick the system. In still another
`embodiment, a plurality of still frame photograms may be captured and
`differences in measurements may be calculated to determine if the
`presented person is an actual person or a photograph.
`
`In a further embodiment, the visitation system 130 may connect the
`requesting parties in an electronic visitation in response to a
`determination that the actual face was presented in the image as shown
`at block 404.
`
`‘420 Patent at 8:61-9:16 (emphasis added).
`
`Verification of an “actual face” is a result of performing the 3D feature
`
`detection process on the captured image. The 3D feature detection is described
`
`having two embodiments (one operating on still frames, the other using video). The
`
`‘420 Patent’s spoof detection distinguishes between an actual person and a
`
`photograph (the actual person having an actual face). The 3D feature detection is
`
`14
`
`
`
`Case IPR2016-01220
`Patent 9,007,420
`described as identifying “three-dimensional characteristics of an image, including
`
`measurements of features of a face at a plurality of points on the image.” ‘420 Patent
`
`at 9:3-6. The 3D feature detection embodiment is clearly a subtype of the feature
`
`detection process. Both 3D feature detection and the feature detection process
`
`implement anti-spoofing by verifying the presence of an actual face, but the
`
`3D feature detection process uses measurements of 3D as opposed to 2D
`
`characteristics of the image. See Kakadiaris Decl., ¶37. Moreover, it is clear from
`
`the context that the verified “actual face” is a face belonging to the “actual person.”
`
`See id. This understanding is also consistent with the process shown in Figure 4
`
`(below).
`
`
`
`15
`
`
`
`Case IPR2016-01220
`Patent 9,007,420
` A Dictionary Definition Provides Insight into the Plain and
`Ordinary Meaning
`
`Webster’s Third New International Dictionary (2002) defines “actual” as
`
`“3: not spurious: REAL, GENUINE.” Ex. 2006 at 22. In the absence of an explicit
`
`definition in the specification, the words of a claim “are generally given their
`
`ordinary and customary meaning.” In re Translogic Tech., Inc., 504 F.3d 1249, 1257
`
`(Fed. Cir. 2007). A general purpose dictionary may provide an ordinary meaning for
`
`a claim term. Google Inc., v. Arendi S.A.R.L., IPR2014-00450, 2015 WL 5016484,
`
`at *7 (Aug. 19, 2015). If a dictionary definition does not contradict any definitions
`
`ascertained by reading of the patent documents, the meaning may be relied upon as
`
`the broadest reasonable interpretation of the term. Id.
`
`This definition is also consistent with the claim language, the specification,
`
`and the figures of the ‘420 Patent. See Kakadiaris Decl., ¶94.
`
` Dr. Kakadiaris provides informed testimony
`Dr. Kakadiaris includes in his declaration his testimony that one having
`
`ordinary skill would have understood “the methods taught in the ‘420 Patent to be
`
`associated with the anti-spoofing technique of distinguishing a live face from a
`
`photograph.” Kakadiaris Decl., ¶90. Dr. Kakadiaris states that a person having
`
`ordinary skill in the art would have understood the methods of measuring the
`
`difference between identified features in an image between frames or at a later point
`
`in a video would allow the detection of an “actual face.” See Kakadiaris Decl., ¶90.
`
`16
`
`
`
`Case IPR2016-01220
`Patent 9,007,420
`Because images of actual people make small, and sometime large, movements
`
`on a constant basis. It is this movement that changes the measured points and their
`
`relative distances between frames. See Kakadiaris Decl., ¶93. This is what the 420
`
`Patent describes. Id.
`
` Prosecution History Disclaimer of the Term’s Scope
`Throughout the prosecution history of the ‘420 Patent, Patent Owner (then
`
`applicant) consistently distinguished prior art and maintained that verifying an actual
`
`face was not disclosed by references that performed “face detection.” For example,
`
`in its first Office Action, the Examiner cited Kaiser (2008/0000966) and another
`
`reference in an obviousness rejection. The Examiner described Kaiser as verifying
`
`an actual face by facial recognition:
`
`Consider claims 1 and 12, Keiser teaches a method and apparatus
`comprising: an image capture device configured to capture an image of
`a user in response to a request to initiate an electronic visitation session
`(par. 0021; 0028-0029); a processing device coupled to the image
`capture device and configured to: perform a feature detection process
`on the image to verify that an actual face was present in the image (par.
`0029; 0039).
`
`History (Ex. 1002) at 68.
`
`Patent Owner responded that detecting, recognizing, and authenticating a
`
`visitor’s image was not “verifying that an actual face of an actual person is present
`
`in the image”:
`
`17
`
`
`
`Case IPR2016-01220
`Patent 9,007,420
`The Office Action asserts that Keiser discloses performing “a feature
`detection process on the image to verify that an actual face was present
`in the image,” citing paragraphs [0029] and [0039] of Keiser. However,
`cited paragraph [0029] of Keiser states, in part: “camera 105 may, in
`block 204, capture an image of the visitor during the registration
`process. Such an image may be stored to registration database 107,
`along with other information captured during the registration process,
`to enable an authorized user to later retrieve such information and verify
`the image of the visitor registering for the visitation session;” “the
`captured image data may be used to compare with, for example, image
`data captured from an identification card to ensure a sufficiently close
`match;” and “the captured image data may be compared with images of
`persons of interest.” Cited paragraph [0039] of Keiser states, in part:
`“controller 108 may compare an image of the visitor captured in block
`204 to stored image information, such as an image scanned from the
`visitor’s identification card, images of persons of interest, etc.;” and
`“controller 108 may determine whether to authorize the requested
`visitation session based at least in part on the comparison of images.”
`
`Hence, Keiser discloses capturing an image and comparing it to another
`image, but fails to teach or suggest verifying that an actual face of an
`actual person is present in an image, as claimed. In, fact Keiser does
`not discuss an “actual face,” or the like in its disclosure, at all.
`
`History at 58-59 (emphasis in original).
`
`The Examiner, like Petitioner here, had confused “actual face” with
`“face.” This was evident is his next response. In response to applicant
`
`18
`
`
`
`Case IPR2016-01220
`Patent 9,007,420
`argument that Keiser does not teach or suggest “verifying that an actual
`face of an actual person is present in an image”. Accordingly, the
`examiner respectfully disagrees. Keiser teaches a visitation and
`registration system for authenticate visitors in a controlled environment
`(abstract; par. 0027). Images of the visitors are captured (via camera)
`and stored during registration process (par. 0028). During visitation
`session, visitor identity is verify via inputs as well as comparing
`captured image with stored image (par. 0029; 0039). Paragraph 0029
`recite: “This may aid in detecting when a visitor posing as another
`person ... Further still in certain embodiments, the captured image data
`may be compared with images of person of interest ....” That is, the
`visitation system captured the visitor image (i.e., actual face of the
`visitor) to verify that the visitor is the actual person that registered by
`comparing with stored data. Therefore, Keiser clearly teach or suggest
`of “verifying that an actual face of an actual person is present in an
`image” as presented in the claims.
`
`History at 47-48 (emphasis in original).
`
`Patent Owner responded by amending claim 12 by expressly stating that the
`
`actual face verification was accomplished “without performing a facial recognition
`
`process.” This is consistent with the Patent Owner’s construction of “actual face”:
`
`Further, with regard to independent claim 12, again assuming, ad
`arguendo, that, as the final Office Action asserts, Keiser discloses “the
`captured image data may be compared with images of person of
`interest,” and/or the like, Keiser clearly fails to teach or suggest
`verifying that an actual face is present in an image by performing a
`
`19
`
`
`
`Case IPR2016-01220
`Patent 9,007,420
`feature detection process on the image, without performing a facial
`recognition process, as now explicitly claimed.
`
`Id., 35, 39.
`
`The claims were subsequently amended by Examiner and allowed as the
`
`claims stand today. Thus, the Prosecution History shows clearly that Patent Owner’s
`
`construction has been consistent and has distinguished mere facial recognition, but
`
`rather covers anti-spoofing.
`
`B. The Construction of “feature detection process” and “three
`dimensional feature detection process” Cannot Ignore and Must be
`Reasonable in Light of the Teachings of the Specification.
`
`Patent Owner construes “three-dimensional (3D) feature detection” to be a
`
`particular type of “feature detection process,” which identifies “three-dimensional
`
`(3D) characteristics of the face present in the image.” See Kakadiaris Decl., ¶98.
`
`Further, “a process for identifying three-dimensional characteristics of the face
`
`present in the image, measuring facial landmarks at a plurality of points on the
`
`image, calculating changes in the measurements in a sequence of video frames or
`
`still images, and determining whether the changes are indicative of an actual face.”
`
`Kakadiaris Decl., ¶101.
`
`Petitioner suggests the term “feature detection process” be construed as “a
`
`process for detecting characteristics of an image, such as measurements of features
`
`of a face at a plurality of points on the image.” Pet. at 23. However, Petitioner ignores
`
`20
`
`
`
`Case IPR2016-01220
`Patent 9,007,420
`the purpose of the “feature detection,” the teachings of the specification, and the
`
`plain meaning of the word “feature” in view of the specification. In each instance
`
`where the term “feature detection process” (or “the detection process”) is recited in
`
`the claims, it refers to the phrase “to verify that an actual face is present.” See ‘420
`
`Patent at 11:9-11, 11:49-50, 12:3-5. Petitioner’s proposed construction of “feature
`
`detection process” ignores this important aspect of the claims and should be rejected.
`
`An appropriate claim construction must include construction of the entire phrase
`
`“performing a [feature] detection process . . . on the [second] image to verify that
`
`[an/the] actual face [was/is] present in the [second] image.” ‘420 Patent at 10:21-24.
`
`The ’420 Patent’s “feature detection process” is included in independent
`
`claims 1 and 11. Each of these claims provide that the feature detection process
`
`operates on a captured image, and is used “to verify that an actual face was present
`
`in the image.” Id. at 1:51-59. The ’420 Patent performs this detection to determine
`
`if “the person is an actual person and not simply a photograph presented to trick the
`
`system.” Id. at 9:5-8. Furthermore, the specification provides that a “captured image
`
`may include either a still frame photograph or a video image.” Id. at 8:55-65. That
`
`is, the captured image may be a “single video frame” or a “plurality of video frames.”
`
`Id. In the disclosed embodiments, the process includes “measurements of features of
`
`a face at a plurality of points on the image.” Id. at 9:2-5.
`
`21
`
`
`
`Case IPR2016-01220
`Patent 9,007,420
`In order to detect whether the captured image is an “actual face” as opposed
`
`to a photograph, embodiments of the process evaluate “changes in the measurements
`
`from frame to frame of a video image” or may use “a plurality of still frame
`
`photograms” that “may be captured and differences in measurements may be
`
`calculated.” Id. at 9:5-12. Thus, changes in the captured image from frame to frame,
`
`such as movement of the lips, eyes, etc. allow the feature detection process “to
`
`determine if the presented person is an actual person or a photograph.” See id.
`
`at 8:66-9:12.
`
`The Board should reject GTL’s overbroad proposed construction because it
`
`ignores this critical aspect of the claimed “feature detection process.” It is
`
`inappropriate to construe the term “feature detection process” in isolation. Instead,
`
`the more complete phrase “performing a [feature] detection process . . . on the
`
`[second] image to verify that [an/the] actual face [was/is] present in the [second]
`
`image” should be construed. Consistent with the plain meaning and in view of the
`
`specification, the claim limitation should be construed as “performing a process
`
`utilizing algorithms for detecting features in a frame and identifying changes or
`
`differences in those features from features in another frame to verify that an actual
`
`face (e.g., not a photograph) is present in the image.”
`
`Claim differentiation does not require otherwise. “Claim differentiation is a
`
`guide, not a rigid rule.” Marine Polymer Techs., Inc. v. Hemcon, Inc., 672 F.3d 1350,
`
`22
`
`
`
`Case IPR2016-01220
`Patent 9,007,420
`1359 (Fed. Cir. 2012). While dependent claims 10 and 19 specifically “utilize[]
`
`three-dimensional (3D) feature detection,” they do not render superfluous the
`
`requirement in independent claim 1 that the “feature detection process . . . verif[ies]
`
`that an actual face was present in the image.” Indeed, throughout the specification,
`
`the “feature detect