`
`__________________
`
`
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`
`
`Mercedes-Benz USA, LLC
`
`Petitioner
`
`
`
`Patent No. 6,772,057
`Issue Date: August 3, 2004
`Title: VEHICULAR MONITORING SYSTEMS USING IMAGE PROCESSING
`
`
`
`
`PETITION FOR INTER PARTES REVIEW
`OF U.S. PATENT NO. 6,772,057
`PURSUANT TO 35 U.S.C. § 312 and 37 C.F.R. § 42.104
`
`Case No. IPR 2014-00646
`
`
`
`
`
`
`
`
`
`
`
`
`
`TABLE OF CONTENTS
`
`LISTING OF EXHIBITS ...................................................................................................... iv
`
`I.
`
`Mandatory Notices (37 C.F.R. § 42.8) ........................................................................ 1
`
`A.
`
`B.
`
`C.
`
`Real Party-in-Interest (37 C.F.R. § 42.8(b)(1)) .............................................. 1
`
`Related Matters (37 C.F.R. § 42.8(b)(2)) ......................................................... 1
`
`Counsel & Service Information (37 C.F.R. §§ 42.8(b)(3)-(4)) ..................... 2
`
`II.
`
`Payment of Fees (37 C.F.R. § 42.103) ........................................................................ 3
`
`III. Requirements For IPR (37 C.F.R. § 42.104) ............................................................. 3
`
`A. Grounds for Standing (37 C.F.R. § 42.104(a)) .............................................. 3
`
`B.
`
`C.
`
`Identification of Challenge (37 C.F.R. § 42.104(b)) and Relief
`Requested (37 C.F.R. § 42.22(a)(1)) ................................................................ 3
`
`Claim Construction (37 C.F.R. § 42.104(b)(3)) ............................................. 5
`
`IV. OVERVIEW of the ’057 Patent ................................................................................ 6
`
`V. How Challenged Claims are Unpatentable (37 C.F.R. §§ 42.104(b)(4)-(5)) ......... 7
`
`A. Ground 1: Claims 1, 2, 4, 7, 16, 40, 41, 43, 46, 56, 59, 60, 61, 77, 78,
`and 81-83 are Anticipated Under 35 U.S.C. § 102(e) by Lemelson ........... 7
`
`B.
`
`C.
`
`Ground 2: Claims 1, 2, 4, 7, 16, 40, 41, 43, 46, 56, 59, 60, 61, 77, 78,
`and 81-83 are Obvious Under 35 U.S.C. § 103(a) in View of
`Lemelson ........................................................................................................... 24
`
`Ground 3: Claims 1, 2, 4, 7, 16, 40, 41, 43, 46, 56, 59, 60, 61, 77, 78,
`and 81-83 are Obvious Under 35 U.S.C. § 103(a) Over Lemelson in
`View of Nishio ................................................................................................. 25
`
`D. Ground 4: Claims 30, 31 and 62 are Obvious Under 35 U.S.C. §
`103(a) Over Lemelson in View of Borcherts ............................................... 28
`
`E. Ground 5: Claims 30, 31 and 62 are Obvious Under 35 U.S.C. §
`103(a) Over Lemelson in View of Komoda ................................................ 30
`
`
`
`ii
`
`
`
`
`
`F.
`
`Ground 6: Claims 30, 31 and 62 are Obvious Under 35 U.S.C. §
`103(a) Over Lemelson in View of Kawai ..................................................... 30
`
`G. Ground 7: Claims 4 and 59 are Obvious Under 35 U.S.C. § 103(a)
`Over Lemelson in View of Asayama ............................................................ 31
`
`H. Ground 8: Claims 43 and 81 are Obvious Under 35 U.S.C. § 103(a)
`Over Lemelson in View of Suzuki ................................................................ 33
`
`I.
`
`J.
`
`Ground 9: Claims 60 and 82 are Obvious Under 35 U.S.C. § 103(a)
`Over Lemelson in View of Ulke .................................................................... 34
`
`Ground 10: Claims 1, 4, 16, 56 and 59 are Anticipated under 35
`U.S.C. § 102(b) by Nishio ............................................................................... 36
`
`K. Ground 11: Claims 2, 4, 40, 41, 43, 59, 77, 78 and 81 are Obvious
`under 35 U.S.C. § 103(a) Over Nishio in View of Asayama ..................... 45
`
`L.
`
`Ground 12: Claims 7 and 61 are Obvious under 35 U.S.C. § 103(a)
`Over Nishio in View of Lemelson ................................................................ 50
`
`M. Ground 13: Claims 30, 31 and 62 are Obvious under 35 U.S.C. §
`103(a) Over Nishio in View of Borcherts .................................................... 51
`
`N. Ground 14: Claims 30, 31 and 62 are Obvious Under 35 U.S.C. §
`103(a) Over Nishio in View of Komoda ...................................................... 53
`
`O. Ground 15: Claims 30, 31 and 62 are Obvious Under 35 U.S.C. §
`103(a) Over Nishio in View of Kawai .......................................................... 53
`
`P.
`
`Ground 16: Claims 40, 43, 77 and 81 are Anticipated Under 35
`U.S.C. § 102(b) by Yamamura........................................................................ 54
`
`Q. Ground 17: Claims 46 and 83 are Obvious Under 35 U.S.C. § 103(a)
`Over Yamamura in View of Lemelson ......................................................... 59
`
`VI. Conclusion .................................................................................................................... 60
`
`
`
`iii
`
`
`
`
`
`Exhibit 1001
`
`Exhibit 1002
`
`Exhibit 1003
`
`Exhibit 1004
`
`Exhibit 1005
`
`Exhibit 1006
`
`Exhibit 1007
`
`Exhibit 1008
`
`Exhibit 1009
`
`Exhibit 1010
`
`Exhibit 1011
`
`Exhibit 1012
`
`Exhibit 1013
`
`LISTING OF EXHIBITS1
`
`U.S. Patent No. 6,772,057 to Breed et al.
`
`U.S. Patent No. 6,553,130 to Lemelson et al.
`
`File History for U.S. Patent Application No. 08/105,304
`
`European Patent Application No. 93112302 (Publication
`No. 0582236A1)
`
`File History for U.S. Patent Application No. 08/097,178
`
`U.S. Patent No. 5,245,422 to Borcherts
`
`Komoda, Norio et al., “Automated Vehicle/Highway
`System,” 13th International Technical Conference on
`Experimental Safety Vehicles, 1991
`
`Kawai, Mitsuo, Collision Avoidance Technologies,
`Leading Change: The Transportation Electronic
`Revolution, 1994
`
`U.S. Patent No. 5,214,408 to Asayama
`
`Suzuki, et al., Driver Environment Recognition for Active
`Safety, Toyota Technical Review Vol. 43, No. 1, Sept.
`1993
`
`Ulke, Walter et al., Radar Based Automotive Obstacle
`Detection System, SAE Technical Paper Series, Feb. 1994
`
`Certified English Translation of Japanese Unexamined
`Patent Application Publication
`JP-H06-124340
`to
`Yamamura
`
`Japanese Unexamined Patent Application Publication
`JP-H06-124340 to Yamamura
`
`
`1 Unless otherwise specified, all citations to Exhibits refer to the original page, column
`
`or line number of that Exhibit.
`
`
`
`iv
`
`
`
`
`
`Exhibit 1014
`
`Exhibit 1015
`
`Exhibit 1016
`
`Exhibit 1017
`
`Exhibit 1018
`
`File History for U.S. Patent Application No. 10/302,105
`
`Infringement Contentions of American Vehicular Sciences
`LLC with respect to U.S. Patent No. 6,772,057 in the
`litigation captioned American Vehicular Sciences LLC v.
`Mercedes-Benz U.S. International, Inc. and Mercedes-Benz USA,
`LLC, 13-cv-00309 (E.D. Tex.)
`
`Expert Declaration of Larry S. Davis
`
`Patent Owner’s March 20, 2014 Response in IPR2013-
`00419
`
`Patent Trial and Appeal Board Decision Instituting Inter
`Partes Review on U.S. Patent No. 6,772,057
`
`
`
`
`
`v
`
`
`
`
`
`Pursuant to 35 U.S.C. §§ 311-319 and 37 C.F.R. Part 42, Mercedes-Benz USA
`
`LLC (“Petitioner”) respectfully requests inter partes review of claims 1, 2, 4, 7, 16, 30,
`
`31, 40, 41, 43, 46, 56, 59-62, 77, 78 and 81-83 of U.S. Patent No. 6,772,057 (“the ’057
`
`patent”). According to U.S. Patent & Trademark Office records, the ’057 patent is
`
`currently assigned to American Vehicular Sciences LLC (“AVS” or “Patent Owner”).
`
`I. MANDATORY NOTICES (37 C.F.R. § 42.8)
`A. Real Party-in-Interest (37 C.F.R. § 42.8(b)(1))
`The real parties-in-interest with respect to this Petition are Petitioner and
`
`Mercedes-Benz U.S. International, Inc. (“MBUSI”).
`
`B. Related Matters (37 C.F.R. § 42.8(b)(2))
`The ’057 patent has been asserted by Patent Owner in the following litigations:
`
`American Vehicular Sciences LLC v. Toyota Motor Corp. et al., 12-CV-410 (E.D. Tex.);
`
`America Vehicular Sciences LLC v. BMW Group et al., 12-CV-415 (E.D. Tex.); American
`
`Vehicular Sciences LLC v. Subaru of Am., Inc., 13- CV-230 (E.D. Tex.); and American
`
`Vehicular Sciences LLC v. Mercedes-Benz U.S. International, Inc. et al., 13-CV-309 (E.D.
`
`Tex.). Petitioner and MBUSI were named as defendants in the 309 Litigation and
`
`served with a Summons and Complaint in that action on April 17, 2013. On July 22,
`
`2013, they were served with infringement contentions in that proceeding. (Ex. 1015,
`
`p. 6.) Pending U.S. Patent App. No. 11/923,929 and numerous other patents and
`
`applications claim the benefit of the application from which the ’057 patent issued.
`
`The ’057 patent is currently the subject of IPR2013-00419 (instituted January 13,
`
`
`
`1
`
`
`
`
`
`2014). Petitioner is not aware of any other pending administrative matter that would
`
`affect, or be affected by, a decision in this proceeding. Petitioner is simultaneously
`
`filing petitions seeking inter partes review of four other patents currently assigned to
`
`AVS: U.S. Patent No. 5,845,000; U.S. Patent No. 6,738,697; U.S. Patent No.
`
`6,746,078; and U.S. Patent No. 7,630,802. These petitions do not address the ’057
`
`patent, but do involve the same patent owner and Petitioner.
`
`C. Counsel & Service Information (37 C.F.R. §§ 42.8(b)(3)-(4))
`Lead Counsel: Scott W. Doyle (Reg. No. 39176)
`
`Back-up Counsel: Jonathan R. DeFosse (pro hac to be requested upon authorization)2.
`
`Electronic Service: scott.doyle@shearman.com, jonathan.defosse@shearman.com.
`
`Postal Address: Scott W. Doyle, Jonathan DeFosse, Shearman & Sterling LLP, 801
`
`Pennsylvania Ave., NW, Suite 900, Washington, DC 20004, 202-508-8000
`
`(telephone), 202-508-8100 (fax).
`
`
`2 Petitioners request authorization to file a motion for Jonathan R. DeFosse to appear
`
`pro hac vice as backup counsel. Mr. DeFosse is an experienced litigation attorney in
`
`patent cases. He is admitted to practice in Virginia and Washington, DC, as well as
`
`before several United States District Courts and Courts of Appeal. Mr. DeFosse is
`
`familiar with the issues raised in this Petition because he represents Petitioners in the
`
`309 Litigation.
`
`
`
`2
`
`
`
`
`
`II.
`
`PAYMENT OF FEES (37 C.F.R. § 42.103)
`
`Petitioner is requesting inter partes review of 21 claims of the ’057 patent. The
`
`United States Patent & Trademark Office is authorized to charge all fees required in
`
`connection with this petition (calculated to be $25,600.00) or these proceedings, to the
`
`deposit account of Shearman & Sterling, LLP, Deposit Account 500324.
`
`III. REQUIREMENTS FOR IPR (37 C.F.R. § 42.104)
`A. Grounds for Standing (37 C.F.R. § 42.104(a))
`Petitioner certifies that the ’057 patent (Ex. 1001) is available for inter partes
`
`review and that Petitioner is not barred or estopped from requesting an inter partes
`
`review challenging the patent claims on the grounds identified in this petition.
`
`B.
`
`Identification of Challenge (37 C.F.R. § 42.104(b)) and Relief
`Requested (37 C.F.R. § 42.22(a)(1))
`
`Petitioner respectfully requests that inter partes review be instituted and claims
`
`1, 2, 4, 7, 16, 30, 31, 40, 41, 43, 46, 56, 59, 60-62, 77, 78 and 81-83 of the ’057 patent
`
`be cancelled on the following grounds of unpatentability:
`
`Ground 1: Claims 1, 2, 4, 7, 16, 40, 41, 43, 46, 56, 59, 60, 61, 77, 78, 81, 82 and
`
`83 are Anticipated Under 35 U.S.C. § 102(e) by Lemelson (Exs. 1002 and 1003).
`
`Ground 2: Claims 1, 2, 4, 7, 16, 40, 41, 43, 46, 56, 59, 60, 61, 77, 78, 81, 82 and
`
`83 are Obvious Under 35 U.S.C. § 103(a) in View of Lemelson
`
`Ground 3: Claims 1, 2, 4, 7, 16, 40, 41, 43, 46, 56, 59, 60, 61, 77, 78, 81, 82 and
`
`83 are Obvious Under 35 U.S.C. § 103(a) Over Lemelson in View of Nishio (Exs.
`
`1004 and 1005).
`
`
`
`3
`
`
`
`
`
`Ground 4: Claims 30, 31 and 62 are Obvious Under 35 U.S.C. § 103(a) Over
`
`Lemelson in View of Borcherts (Ex. 1006).
`
`Ground 5: Claims 30, 31 and 62 are Obvious Under 35 U.S.C. § 103(a) Over
`
`Lemelson in View of Komoda (Ex. 1007).
`
`Ground 6: Claims 30, 31 and 62 are Obvious Under 35 U.S.C. § 103(a) Over
`
`Lemelson in View of Kawai (Ex. 1008).
`
`Ground 7: Claims 4 and 59 are Obvious Under 35 U.S.C. § 103(a) Over
`
`Lemelson in View of Asayama (Ex. 1009).
`
`Ground 8: Claims 43 and 81 are Obvious Under 35 U.S.C. § 103(a) Over
`
`Lemelson in View of Suzuki (Ex. 1010).
`
`Ground 9: Claims 60 and 82 are Obvious Under 35 U.S.C. § 103(a) Over
`
`Lemelson in View of Ulke (Ex. 1011).
`
`Ground 10: Claims 1, 4, 16, 56 and 59 are Anticipated under 35 U.S.C. § 102(b)
`
`by Nishio
`
`Ground 11: Claims 2, 4, 40, 41, 43, 59, 77, 78 and 81 are Obvious under 35
`
`U.S.C. § 103(a) Over Nishio in View of Asayama
`
`Ground 12: Claims 7 and 61 are Obvious under 35 U.S.C. § 103(a) Over
`
`Nishio in View of Lemelson
`
`Ground 13: Claims 30, 31 and 62 are Obvious under 35 U.S.C. § 103(a) Over
`
`Nishio in View of Borcherts
`
`
`
`4
`
`
`
`
`
`Ground 14: Claims 30, 31 and 62 are Obvious Under 35 U.S.C. § 103(a) Over
`
`Nishio in View of Komoda
`
`Ground 15: Claims 30, 31 and 62 are Obvious Under 35 U.S.C. § 103(a) Over
`
`Nishio in View of Kawai
`
`Ground 16: Claims 40, 43, 77 and 81 are Anticipated Under 35 U.S.C. § 102(b)
`
`by Yamamura (Exs. 1012 and 1013).
`
`Ground 17: Claims 46 and 83 are Obvious Under 35 U.S.C. § 103(a) Over
`
`Yamamura in View of Lemelson
`
`The above-listed grounds of unpatentability are explained in detail in Section
`
`[V], below. This Petition is supported by the Declaration of Larry S. Davis (Ex. 1016).
`
`C. Claim Construction (37 C.F.R. § 42.104(b)(3))
`The Board construed certain terms of the ’057 patent in its Decisions granting
`
`the Toyota petitions for inter partes review. (Ex. 1018, pp. 7-15.) The below chart has
`
`a summary of those constructions, which Petitioner adopts3 in this petition.
`
`Claim Term
`“trained pattern
`
`Board’s Construction
`“an algorithm that processes a signal that is generated by
`
`
`3 In the 309 Litigation, Petitioner has taken the position that the following terms of
`
`the ’057 patent are indefinite under 35 U.S.C. § 112(b): “trained pattern recognition
`
`means…” (claims 1, 16, 31, 41, 56). Petitioner has no opportunity to challenge this
`
`term as indefinite under § 112(b) as part of the IPR proceedings.
`
`
`
`5
`
`
`
`
`
`recognition
`
`an object, or is modified by interacting with an object, in
`
`algorithm” (claims 1,
`
`order to determine to which one of a set of classes the
`
`31, 41, and 56)
`
`object belongs, the algorithm having been taught, through
`
`a variety of examples, various patterns of received signals
`
`generated or modified by objects” (Id. at 8.)
`
`“a neural computer or neural network trained for pattern
`
`recognition, and equivalents thereof” (Id. at 9-12.)
`
`“determine that the object belongs to a particular set or
`
`class” (Id. at 12.)
`
`“trained pattern
`
`recognition means…”
`
`(1, 16, 31, 41 and 56)
`
`“identify” (ubiquitous)
`
`“exterior object”
`
`“a material or physical thing outside the vehicle, not a part
`
`(ubiquitous)
`
`of the roadway on which the vehicle travels” (Id. at 12-14.)
`
`“rear view mirror”
`
`“a mirror that faces to the rear, which necessarily excludes
`
`(30, 62)
`
`non-rear-facing mirrors” (Id. at 14.)
`
`“transmitter” (4, 43,
`
`“devices that transmit any type of electromagnetic waves,
`
`59 and 81)
`
`including visible light.” (Id. at 14-15.)
`
`IV. OVERVIEW OF THE ’057 PATENT
`The ’057 patent generally relates to a vehicle monitoring system that utilizes
`
`various types of sensors such as cameras, radar or laser radar (lidar) in order to detect
`
`objects. (Ex. 1001, 17:53-23:9, 39:1-20.) In one embodiment, a processor receives the
`
`
`
`6
`
`
`
`
`
`signals from the sensor(s) and identifies, classifies or locates an object using a trained
`
`pattern recognition algorithm. (Id. at 14:8-25, 8:15-19.) A system in the vehicle, such
`
`as a visual display, can then be affected depending on the classification, identification
`
`or location of the exterior object. (Ex. 1001, 14:26-28; Ex. 1014, p. 16.)
`
`Petitioner challenges five independent claims, claims 1, 30, 40, 56 and 77.
`
`Independent claims 1 and 56 are very similar, and require (i) a “at least one receiver”
`
`arranged to receive waves from the vehicle exterior, (ii) a “processor comprising
`
`trained pattern recognition means” that applies a “trained pattern recognition
`
`algorithm” to provide the “classification, identification or location of the exterior
`
`object” and (iii) a system in the vehicle that is “affected in response” thereto. Claim
`
`16 requires (i) a “CCD array,” (ii) “a processor… comprising trained pattern
`
`recognition means,” to provide the “classification, identification or location of the
`
`exterior object” and (iii) a system in the vehicle that is “affected in response” thereto.
`
`Independent claim 30 is similar to claims 1 and 56 but requires the “at least one
`
`receiver” to be “arranged on a rear view mirror of the vehicle” and does not require
`
`“trained pattern recognition means” or a “trained pattern recognition algorithm.”
`
`Independent claims 40 and 77 require a “plurality of receivers,” but do not require
`
`“trained pattern recognition means” or a “trained pattern recognition algorithm.”
`
`V. HOW CHALLENGED CLAIMS ARE UNPATENTABLE (37 C.F.R.
`§§ 42.104(B)(4)-(5))
`A. Ground 1: Claims 1, 2, 4, 7, 16, 40, 41, 43, 46, 56, 59, 60, 61, 77, 78,
`and 81-83 are Anticipated Under 35 U.S.C. § 102(e) by Lemelson
`
`
`
`7
`
`
`
`
`
`Claims 1, 2, 4, 7, 16, 40, 41, 43, 46, 56, 59, 60, 61, 77, 78, and 81-83 are
`
`anticipated by Lemelson under 35 U.S.C. § 102(e). Lemelson describes a vehicle
`
`exterior monitoring system that one of ordinary skill could implement to identify
`
`exterior objects and obstacles and affects a vehicle system by warning the driver by,
`
`for example, actuating the brakes or steering to minimize the likelihood or effects of a
`
`collision. (Ex. 1002, Abstract; 2:53-63, 3:5-26, 5:15-18, 8:38-39, Fig. 1; Ex. 1003, pp. 1,
`
`3-6, 8, 13-14, Fig. 1; Ex. 1016, ¶ 22.) Figure 1 of Lemelson depicts a radar/lidar
`
`computer 14 for locating an exterior object based on received radar or lidar signals, a
`
`camera receiver 16 to receive waves from the exterior environment, an image analysis
`
`computer 19 (“IAC”) for classifying and identifying exterior objects, brakes 33 and
`
`steering 36 that are affected depending on the identified exterior objects and a display
`
`32 for warning the driver of a potential collision. (Ex. 1002, 5:31-56, 5:67-6:8; Ex.
`
`1003, pp. 8-9.) Lemelson teaches that the signal output from the camera(s) is digitized
`
`and passed to the IAC. (Ex. 1002, 5:36-39; Ex. 1003, pp. 8-9.) The IAC “identifies”
`
`the detected exterior object(s) using “neural networks” that have been “trained” using
`
`“known inputs.” (Ex. 1002, 5:39-45, 7:47-8:10, 8:21-23; Ex. 1003, pp. 9, 12-13.)
`
`Lemelson anticipates independent claims 1, 16, 40, 56 and 77. First, Lemelson
`
`teaches the “receiver…” limitations (“plurality of receivers” in claims 40 and 77) of
`
`these claims because it teaches several devices that are coupled to a processor and
`
`receive electromagnetic radiation. (Ex. 1002, Figs. 1, 2, 6:21-42; Ex. 1003, Figs. 1, 2,
`
`p. 10) For example, Lemelson discloses radar and lidar receivers, (Ex. 1002, 5:67-6:8;
`
`
`
`8
`
`
`
`
`
`
`
`Ex. 1003, p. 13), as well as “multiple cameras” that are used “for stereo imaging
`
`capabilities” (Ex. 1002, Figs. 1 and 2, 6:37-38; Ex. 1003, Figs. 1 and 2, p. 9; Ex. 1016,
`
`¶¶ 25-27.) Lemelson also teaches use of “a CCD array camera” as a receiver. (Ex.
`
`1002, 6:28-34, 7:36-41; Ex. 1003, pp. 10, 12.) Claims 40 and 77 require “a plurality of
`
`receivers arranged apart from one another and to receive waves from different parts
`
`of the environment exterior of the vehicle…” Lemelson meets this limitation at least
`
`through its disclosure that these “[m]ultiple cameras may be used for front, side and
`
`rear viewing and for stereo imaging capabilities suitable for generation of three
`
`dimensional image information including capabilities for depth perception and
`
`placing multiple objects in three dimensional image fields to further improve hazard
`
`detection capabilities.” (Ex. 1002, 6:37-42; Ex. 1003, p. 10).
`
`Second, Lemelson teaches processing the received signals to provide a
`
`classification, identification or location of the exterior object. For example, Lemelson
`
`teaches that “[t]he computer is operable to analyze video and/or other forms of
`
`image information generated as the vehicle travels to identify obstacles ahead of the
`
`vehicle…” (Ex. 1002, 2:39-41; Ex. 1003, p. 3.) Lemelson teaches that the analog
`
`signal output from the cam era(s) is passed to the IAC. (Ex. 1002, 5:36-39; Ex. 1003,
`
`pp. 8-9.) Further, the IAC meets the “trained pattern recognition means…”
`
`limitation as construed by the Board because it discloses a processor that implements
`
`a “neural network” and is therefore a “neural computer.” (Ex. 1016, ¶¶ 28-32.)
`
`“Neural networks” are defined by the ’057 patent to be a type of “trained pattern
`
`9
`
`
`
`
`
`
`
`recognition algorithm” and are within the Board’s construction of that term. (Ex.
`
`1001, 4:35-36.) In particular, Lemelson discloses that the IAC is “provided,
`
`implemented and programmed using neural networks and artificial intelligence as well
`
`as fuzzy logic algorithms” to “(a) identify objects on the road ahead such as other
`
`vehicles, pedestrians, barriers and dividers, turns in the road, signs and symbols, etc.,
`
`and generate identification codes, and (b) detect distances from such objects by their
`
`size (and shape) . . . .” (Ex. 1002, 5:39-45; Ex. 1003, p. 9.) Lemelson explains that
`
`the neural network in the IAC may be “trained” using “known inputs.” (Ex. 1002,
`
`7:47-8:10, 8:21-23; Ex. 1003, pp. 12-13.) This disclosure meets both the “trained
`
`pattern recognition means…” limitation of claims 1, 16 and 56 as construed by the
`
`Board, and the broader “classify, identify or locate” limitation of claim 40 and 77.
`
`The IAC, which uses a neural network, “classifies, identifies and locates”
`
`exterior objects. Neural networks are inherently classifiers because they attribute a
`
`label to a given input value—i.e. determine that an object belongs to a particular set or
`
`class. (Ex. 1016, ¶¶ 29-32.) The IAC “classifies” by using a neural network program
`
`that is “structured and arranged to apply a trained pattern recognition algorithm
`
`generated from data of possible exterior objects and patterns of received waves”
`
`therefrom. AVS argues that Lemelson does not teach this limitation because it does
`
`not teach an “algorithm generated from data of possible exterior objects and patterns
`
`of received waves” therefrom. (Ex. 1017, p. 11.) This is incorrect. To the contrary,
`
`the broadest reasonable construction of this limitation does not require that the
`
`10
`
`
`
`
`
`
`
`training set be directly imaged from physical exterior objects. Nowhere do the
`
`challenged claims require, or even imply, that such data must be imaged directly from
`
`actual exterior objects. AVS imports this limitation into its claims in a desperate
`
`attempt to save them from anticipatory prior art. The disputed term could mean any
`
`type of data so long as it relates to information about such objects, irrespective of
`
`whether it is real image data or synthetically generated. (Id. at ¶¶ 33.)
`
`In any event, even if the disputed limitation were construed by the Board to
`
`require data imaged directly from actual exterior objects, Lemelson teaches this. In
`
`particular, Lemelson teaches that “training involves providing known inputs to the
`
`network” and that “[a]daptive operation is also possible with on-line adjustment…”
`
`(Ex. 1002, 8:4-10; Ex. 1003, p. 13.) This disclosure would necessarily convey to one
`
`of ordinary skill in the art that the neural network of Lemelson was trained on images
`
`directly obtained from actual objects (i.e. natural image data). “Adaptive operation”
`
`could only have been accomplished with direct imaging using onboard vehicle
`
`sensors, which indicates that the “known inputs” Lemelson refers to mean natural
`
`image data. (Ex. 1016, ¶¶ 34-38.)
`
`In the early-to-mid 1990s, one of ordinary skill would have known that the
`
`statistical patterns provided by real imagery—essential in training a neural network to
`
`recognize complex 3-dimensional objects such as the “automobiles, trucks, and
`
`pedestrians” mentioned in Lemelson—could not have been found in synthetic data
`
`(and, in many cases, still cannot be found in such data today). (Id. at ¶¶ 33-36.) In
`
`11
`
`
`
`
`
`
`
`the absence of the statistical patterns present in natural image data, a neural network
`
`will not learn to recognize real objects such as automobiles, trucks and pedestrians as
`
`is necessary in the field of vehicle safety. (Id.)
`
`In the early-to-mid 1990s, directly imaged data was by far the most realistic
`
`type of data that could be obtained to train a neural network to classify or identify
`
`the virtually limitless variety of complex 3-dimensional objects a vehicle would be
`
`expected to encounter in operation. (Id. at ¶¶ 34-36.) One of ordinary skill would
`
`not have expected to succeed in training a network for this task without using
`
`sufficiently representative images of objects the vehicle would be expected to
`
`encounter. (Id.) Such sufficiently representative images could only have been
`
`obtained from real image data. (Id.)
`
`Indeed, it is very telling Lemelson provides no disclosure whatsoever as to the
`
`complex 3-dimensional models necessary to produce such data synthetically. Such
`
`models would necessarily have had to render both the surfaces and reflectance
`
`patterns of complex objects like human beings, requiring enormous processing
`
`power that was prohibitively expensive and generally unavailable in the early-to-mid
`
`1990s. (Id. at ¶ 34-36.) Moreover, such models would still not have produced an
`
`acceptable substitute for natural image data. During this period, synthetic data was
`
`vastly inferior to real image data (and still is in many cases), and required expensive,
`
`time-consuming and complex models to produce. That Lemelson says nothing
`
`12
`
`
`
`
`
`
`
`about how such a synthetic model was developed or implemented, indicates that
`
`Lemelson’s disclosure relates to real image data. (Id.)
`
`Moreover, Lemelson itself cites to references which explicitly disclose training
`
`a neural network on real data. (Id. at ¶ 38.) For example, Lemelson cites to an article
`
`entitled “Integration of Acoustic and Visual Speech Signals Using Neural Networks”
`
`(“Yuhas.”) (Ex. 1002, 19:26-28; Ex. 1003, p. 56.) Yuhas teaches training a neural
`
`network on visual and auditory speech data in order to enable the network to
`
`recognize human speech. Yuhas, B. P., et. al, Integration of Acoustic and Visual Speech
`
`Signals Using Neural Networks, IEEE Communications Magazine, pp. 65-71, Nov.,
`
`1989. In particular, Yuhas teaches that “speech signals… were obtained from a male
`
`speaker who was videotaped…” and that the resulting images were “sampled to
`
`produce a topographically accurate image of 20 x 25 pixels.” (Id. at 67.) As Yuhas
`
`exemplifies, one of ordinary skill would necessarily have resorted to training a neural
`
`network on directly-imaged data to teach it to recognize (i.e. accurately ascribe labels
`
`to) complex objects, such as a human face. (Ex. 1016, ¶¶ 34-38.)
`
`Additionally, the argument that training only on a specific feature of an object
`
`(such as the nose on a person’s face) is not training a system with “data of possible
`
`exterior objects” is meritless. Again, this is an attempt to distinguish the claims from
`
`anticipatory prior art by importing the requirement that the “whole” object must be
`
`trained on. There is no support in the plain language of the claims for this assertion.
`
`Training on a specific feature of an object is training on the object itself because the
`
`13
`
`
`
`
`
`information received from and about the specific feature would still originate from
`
`or be generated by the object. (Id. at ¶ 38.)
`
`Finally, Lemelson discloses the limitations of claims 1, 16, 40, 56 and 77
`
`requiring that a vehicle system be affected in response to the classification,
`
`identification or location of the exterior object. (Id. at ¶¶ 39, 40.) Based on the neural
`
`network object determination, the IAC can display “symbols representing the hazard
`
`objects.” (Ex. 1002, 6:43-55, 9:60-62, Fig. 2; Ex. 1003, pp. 10-11, 16.) The IAC also
`
`provides codes to a decision computer 23, which “integrates the inputs from the
`
`image analysis computer 19” as well as a “radar or lidar computer 14.” (Ex. 1002,
`
`8:30-33, 6:1-8; Ex. 1003, pp. 9, 13.) The decision computer 23 then generates control
`
`signals to control a vehicle system such as brakes or steering. (Ex. 1002, 5:46-51,
`
`2:53-3:26; Ex. 1003, pp. 4, 9.) Thus, Lemelson anticipates each of independent claims
`
`1, 16, 40, 56 and 77. (Ex. 1016, ¶¶ 41, 46-49.)
`
`Lemelson also anticipates claims 2, 4, 7, 41, 46, 59, 61, 78, and 83. Claim 2
`
`depends directly from claim 1 but additionally requires that the “at least one receiver
`
`comprises a pair of receivers spaced apart from one another.” Lemelson teaches that
`
`“[m]ultiple cameras may be used” for “stereo imaging capabilities.” (Ex. 1002, 6:37-
`
`38; Ex. 1003, p. 10.) Receivers (such as cameras) spaced apart from one another are
`
`necessary for such “stereo imaging capabilities.” (Ex. 1016, ¶ 42.)
`
`
`
`14
`
`
`
`
`
`Claims 4 (depends from 1) and 59 (depends from 56) require a “transmitter for
`
`transmitting waves into the environment exterior of the vehicle whereby the at least
`
`one receiver” that is “arranged to receive waves transmitted by said transmitter and
`
`reflected by any exterior objects.” Lemelson meets this because it teaches a vehicle
`
`with “headlights,” (Ex. 1002, 3:29, 5:57; Ex. 1003, pp. 5, 9), which project light that is
`
`reflected off exterior objects. Vehicle headlights are a “transmitter” within the
`
`Board’s construction because they “transmit… electromagnetic waves.” (Ex. 1016,
`
`¶¶ 43, 54.) Lemelson teaches that a camera (or cameras, for stereo imaging
`
`capabilities) may be used for “front” viewing. (Ex. 1002, 6:37-38; Ex. 1003, p. 10.)
`
`Such a positioned camera (or cameras) would necessarily receive light transmitted by
`
`the vehicle headlights and reflected off exterior objects. (Ex. 1016, ¶ 43, 44.)
`
`Claims 7 (depends directly from 1), 46 (depends directly from 40), 61 (depends
`
`directly from 56) and 83 (depends directly from 77) require that the affected vehicle
`
`system is a “display viewable to the driver and arranged to show an image or icon of
`
`the exterior object.” Lemelson teaches that the IAC can display “symbols
`
`representing the hazard objects.” (Ex. 1002, 6:43-55, Fig. 2, 9:60-62; Ex. 1003, pp. 10-
`
`11, 16, Fig. 2; Ex. 1016, ¶ 45.) Lemelson also teaches “various warning and vehicle
`
`operating devices such as… a display driver 31 which drives a (heads-up or
`
`dashboard) display.” (Ex. 1002, 5:49-56; Ex. 1003, p. 9.) Thus, Lemelson anticipates
`
`claims 7, 46, 61 and 83. (Ex. 1016, ¶¶ 45, 53, 56, 60.) Claims 41 (from 40) and 78 (
`
`
`
`15
`
`
`
`
`
`from 77) include the same “trained pattern recognition means” limitation found in
`
`claims 1 and 56 and are therefore also anticipated by Lemelson. (Ex. 1016, ¶¶ 50, 57.)
`
`Claims 43 and 81 each recite the additional limitation “wherein the monitoring
`
`arrangement further comprises a transmitter for transmitting waves into the
`
`environment exterior of the vehicle whereby said receivers are arranged to receive
`
`waves transmitted by said transmitter and reflected by any exterior objects.” And, by
`
`virtue of their dependence from claims 40 and 77, respectively, each also requires “a
`
`plurality of receivers arranged apart from one another and to receive waves from
`
`different parts of the environment exterior of the vehicle.” However, these claims do
`
`not require any “trained pattern recognition means” because independent claims 40
`
`and 77 only require “a processor coupled to said receivers and arranged to classify,
`
`identify or locate the exterior object,” without specifying the manner by which such
`
`classification, identification or location is to occur.
`
`Lemelson anticipates claims 43 and 81 at least through its disclosure that “radar
`
`or lidar scanning may be jointly employed to identify and indicate distances
`
`between the controlled vehicle and objects ahead of, to the side(s) of, and to the
`
`rear of the controlled vehicle.” Emphasis added. (Ex. 1002, 6:5-9; Ex