`Trials@uspto.gov
`
`571-272-7822
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`____________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`____________
`
`APPLE, INC.,
`Petitioner,
`
`v.
`
`COREPHOTONICS LTD.,
`Patent Owner.
`____________
`
`IPR2020-00861 (Patent 10,230,898 B2)
`IPR2020-00862 (Patent 10,356,332 B2)1
`____________
`
`Record of Oral Hearing
`Held Virtually: Thursday, September 9, 2021
`____________
`
`
`
`
`
`
`
`
`Before BRYAN F. MOORE, MONICA S. ULLAGADDI, and
`BRENT M. DOUGAL, Administrative Patent Judges.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`IPR2020-00861 (Patent 10,230,898 B2)
`IPR2020-00862 (Patent 10,356,332 B2)1
`
`
`
`APPEARANCES:
`
`ON BEHALF OF THE PETITIONER:
`
`
`DAVID O'BRIEN, ESQUIRE
`HAYNES & BOONE, LLP
`600 Congress Avenue
`Suite 1300
`Austin, TX 78701
`(512) 867-8400
`
`
`
`ON BEHALF OF THE PATENT OWNER:
`
`
`MARC FENSTER, ESQUIRE
`JONATHAN LINK, ESQUIRE
`RUSS AUGUST & KABAT
`12424 Wilshire Boulevard
`12th Floor
`Los Angeles, California 90025
`(310) 826-7474
`
`
`
`
`
`The above-entitled matter came on for hearing on Thursday, September
`9, 2021, commencing at 9:01 a.m. EST, by video/by telephone.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`2
`
`
`
`
`IPR2020-00861 (Patent 10,230,898 B2)
`IPR2020-00862 (Patent 10,356,332 B2)1
`
`P R O C E E D I N G S
` JUDGE MOORE: Good afternoon. We are here today
`for oral arguments in Inter Partes Review numbers 2020-00861
`and 862. Let's start the hearing by getting appearances by
`the counsel. Who is here on behalf of the petitioner?
` MR. O'BRIEN: Thank you, Your Honor. This is David
`O'Brien at Haynes & Boone in Austin, Texas for petitioner
`Apple. I'm also here with Hong Shi, my colleague. And I
`believe that Aaron Huang from Apple and Priya Viswanath from
`Cooley are both on the line as well.
` JUDGE MOORE: All right. Thank you. And then for
`patent owner?
` MR. FENSTER: Good morning, Your Honors. This is
`Marc Fenster with Russ August & Kabat on behalf of patent
`owner, Corephotonics. With me on the line are my colleagues
`Neil Rubin, Jonathan Link, and James Tsuei.
` JUDGE MOORE: All right. Thank you. With me this
`morning -- I'm Judge Moore. We have Judge Ullagaddi and Judge
`Dougal. So as you're aware, this hearing is being held
`remotely by video conference. We want to make sure everyone
`is going to be heard. If at any time during the hearing you
`have technical or other difficulties that you feel undermine
`your ability to adequately represent your client please let us
`know immediately by contacting the team member who provided
`you with the connection information. We will address any
`issues that arise. If necessary, we can adjust the time to
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`3
`
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`
`IPR2020-00861 (Patent 10,230,898 B2)
`IPR2020-00862 (Patent 10,356,332 B2)1
`account for these technical issues.
` The judges have the parties' demonstratives but
`please remember when referring to the demonstratives to
`identify the slide you're on so that we can follow along.
`Also, if you refer to the record, make sure that you refer to
`where in the record, exhibit number, et cetera, and give the
`judges time to catch up with you and pull up the record that
`you're referring to.
` Please identify yourself when speaking so that the
`reporter can accurately record the transcript. I will also
`remind you that the recording of this proceeding either by
`audio or video is prohibited.
` As set forth in the oral hearing order, each party
`will have 90 minutes total time to argue the two cases.
`Petitioner, who bears the burden to show unpatentability, will
`begin by presenting its case-in-chief and then patent owner
`will respond thereafter the petitioner. You may reserve time
`for rebuttal and the patent owner may reserve time as well.
`You may not reserve more than half your time for rebuttal.
` I would also remind the parties that any objection,
`unless they go to confidentiality or some issue where a
`confidential matter would be revealed, but any other objection
`should be held and not interrupt the other side's
`presentation. Give us those rebuttals during your
`presentation.
` I want to say I think in these cases we do have
`secondary considerations. I believe these parties and this
`4
`
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`
`IPR2020-00861 (Patent 10,230,898 B2)
`IPR2020-00862 (Patent 10,356,332 B2)1
`counsel have dealt with this before in these cases. Can
`counsel let me know if there are issues of confidentiality or
`if that is going to be something that we will be talking about
`today? And I guess I'll start with petitioner.
` MR. O'BRIEN: Subject to needing to respond to
`something from the patent owner, none of our material will be
`confidential. I believe there are slides identified by the
`petitioner as subject to the protective order.
` JUDGE MOORE: Okay. Well, I think what you're
`telling me is that you don't plan on entering into any
`confidential information. If you do, let me know so that we
`can act accordingly. And patent owner?
` MR. FENSTER: This is Marc Fenster. We do not
`anticipate introducing or discussing confidential information
`at today's hearing.
` JUDGE MOORE: All right. Thank you. The other
`thing I saw that patent owner had two sets of slides. My
`sense in these cases are that they overlap very heavily. I
`think it would make sense for these cases -- to discuss them
`together. So if that is going to be a problem for patent
`owner we could certainly give you some time to adjust to that,
`but assuming that that's okay, we will go and do the full 90
`minutes for each party.
` So I want to have petitioner come up. Tell me how
`much time you would like to reserve total for both cases and
`we'll proceed on that basis. Again, I think I may have
`misspoke. I think petitioner also has two sets but that's I
`5
`
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`
`IPR2020-00861 (Patent 10,230,898 B2)
`IPR2020-00862 (Patent 10,356,332 B2)1
`think the best way to proceed. Again, if the parties need
`some time or that causes any disruption please let me know.
`We'll see that it does not. We'll start with petitioner.
` MR. O'BRIEN: No problem with petitioner, Your
`Honor. In fact, you anticipated two of our questions. So I
`understand that we're going to go the full 90 for the two
`proceedings given the commonality of the subject matter. So
`that answers the consolidated transcript question. That's
`great. Will we be going with the order that we have in some
`of the prior proceedings of patent owner going forward first
`on the issues on which they bear the burden, namely the
`secondary factors?
` JUDGE MOORE: Yes.
` MR. O'BRIEN: Or will this all be a single
`integrated hearing?
` JUDGE MOORE: We can go forward if -- I got the
`sense of what I heard from patent owner that we wouldn't be
`covering that, but I'll turn to them and see if that's
`something we need to deal with. Patent owner are you going to
`give a presentation on secondary considerations?
` MR. FENSTER: Your Honor, this is Marc Fenster. So
`we had anticipated arguing pursuant to the scheduling order.
`That I would be arguing the 861 and my partner, Jonathan Link,
`would be arguing the 862. So we had anticipated two attorneys
`splitting the argument. I think that we can accommodate Your
`Honor's suggestion to have a consolidated argument doing
`everything at once. What I would suggest is that I'll start
`6
`
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`
`IPR2020-00861 (Patent 10,230,898 B2)
`IPR2020-00862 (Patent 10,356,332 B2)1
`and then turn it over to Mr. Link in our case-in-chief to
`address any issues specific to the 862. And we may split the
`surrebuttal as well. So if that is acceptable to Your Honor
`then we are okay having the -- you know, combining the two.
`And as for the secondary considerations, I think that we'll be
`addressing the merits and leaving the secondary -- you know, I
`think we'll rest on briefs with respect to the secondary
`conditions. We're certainly not waiving those but don't think
`they merit further discussion here.
` JUDGE MOORE: Certainly. Okay. So I think that
`resolves all the preliminary issues. We don't have any issues
`with changing the order regarding secondary considerations.
`So with that, unless there's anything else from the other
`judges, petitioner just let me know how much time you would
`like to reserve once again for -- out of your total 90
`minutes.
` MR. O'BRIEN: I'll reserve 30 minutes. 60 should
`be ample.
` JUDGE MOORE: Okay. 30 minutes. All right. Okay.
`So whenever you're ready.
` MR. O'BRIEN: Okay. So thank you Your Honors. May
`it please the board. I am going to speak from my slide deck
`that is for the 861 proceeding, so the 898 patent. You have
`already identified that there is a significant amount of
`commonality between these two and in fact the slide decks are
`virtually identical for both partes. The differences between
`the slide decks in terms of your using them as a reference
`7
`
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`
`IPR2020-00861 (Patent 10,230,898 B2)
`IPR2020-00862 (Patent 10,356,332 B2)1
`into the record, the citations of course are particular to the
`patent. So I'm going to be speaking to the 861 slides as my
`backbone for the presentation and then I'll jump on points
`that are particular to the 862 which are frankly few and far
`between.
` So starting with slide 2 and sort of reiterating
`that point, you'll see the commonality between the aspects of
`in each case claim 1. Also, the other independent claim in
`each of the (indiscernible). The primary point of variation
`is just in a small switch up in how the -- either the up-
`transfer ZF or the ZFT is expressed. They are essentially the
`same concept with perhaps a little more wiggle room around the
`up-transfer ZF than around the ZFT which is defined as a ratio
`of tangents. But again, I'll be starting with the 898 and the
`terminology in the 898 is the up-transfer ZF.
` Slide 3 identifies the grounds. You're I'm sure
`familiar with those. You're also familiar with the issues and
`what items are undisputed. Let me go first to -- next to
`slide 4 and that is claim 1. First off, let me apologize. We
`have a slight correction on this slide. The yellow and green
`symbology does not precisely track with our identification of
`1.4 and 1.5. The yellow should continue down through the
`phrase "includes wide image data,". So just visually
`understand that the swatch of what we have identified in the
`petition and treat in our papers as 1.4 extends down another
`clause and that the green portion for 1.5 is actually a little
`shorter than the slide is suggesting.
`8
`
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`
`IPR2020-00861 (Patent 10,230,898 B2)
`IPR2020-00862 (Patent 10,356,332 B2)1
` But at its essence, claim 1, it should be very
`familiar to Your Honors. It recites the very well-known
`technique of using two lenses, each with a fixed field of view
`and digitally zooming through those lenses to provide an
`extended zoom range. This is well known in both digital
`cameras, in mobile devices, at varieties of scale, this
`technique.
` And if you look at the claim, it recites the two
`cameras. It recites, as we've demonstrated before with sort
`of hand visuals, the wide lens or wide field of view lens, and
`a tele lens with a narrower field of view. It speaks to the
`camera controller that provides the zoom-in operation. So we
`zoom from the wide using the wide lens and image sensor. We
`zoom in to a point at or after which we are able to switch to
`the tele. And that switch point, that at or after point, is
`the point at or after which the zoomed in portion of the wide
`is fully subsumed within the field of view of the tele. At
`that point it's very easy to switch and we can continue the
`zooming process with the tele.
` Again, this should be both well known to the board
`and well known in the art. There are many references that
`teach this and we will discuss several of them here. At the
`at or after, at that switch point, and in this claim it said
`it involves the up transfer for ZF. The system switches to
`the tele and it switches to a transformed digitally zoomed
`tele. So switching to that transformed version. The
`transformation has to do with aligning to the other one.
`9
`
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`
`IPR2020-00861 (Patent 10,230,898 B2)
`IPR2020-00862 (Patent 10,356,332 B2)1
`Again, this is that jump effect that we've spoken to in many
`of the other cases. What this, and frankly each of the pieces
`of prior art that we're going to discuss, the problem that
`they're each trying to solve is when we transition from the
`wide to the tele we don't want -- if I am the subject of the
`motion video, we don't want me to jump to the side because of
`a parallax shift between the images. So that's the
`transformation that it speaks to. Again, all very familiar I
`think to Your Honors, familiar to a person of skill in the
`art.
` The inventive step that this patent, the 898, and
`also the 332 purports to add is the addition of what they call
`a no-switching criterion and that encapsulates really the
`common sense insight of if my scheme is to transition or
`transform to switch from wide to the tele, if there are
`situations in which switching from the wide to the tele would
`not provide a benefit, it might degrade the image in some way,
`maybe I might not want to do that. So the non-switching
`criterion is about saying let's not do that under certain
`circumstances.
` So that's the overlay of this claim and you'll see
`what this claim adds then is two forks. It adds the -- it
`includes what we first discussed which was the normal
`operation which is to go to the transformed tele and that's
`really in limitation 1.5. 1.4 says -- and that's what happens
`when the non-switch or the no-switching criterion is not. And
`I apologize for the double negatives. We didn't write the
`10
`
`
`
`
`IPR2020-00861 (Patent 10,230,898 B2)
`IPR2020-00862 (Patent 10,356,332 B2)1
`claims, but that's how it works.
` And in the other situation where the no-switching
`criterion is fulfilled, so don't switch, that's the portion
`that's in the 1.4 portion and it says stick with the wide. So
`continue doing your zoom in on the wide. Don't do what you
`would normally do which is to switch to the tele. And the
`reason that you would not do that is because it wouldn't be a
`good idea. It might result in a degraded image, blurry image,
`something of that sort. So that's what this claim is about.
`That is also what the 332 claim is about. You know, again,
`it's just a slightly different formulation of the up-transfer
`ZF or the ZFT switch point.
` So moving forward then to slide 5. In this case
`and in both cases quite frankly there is no claim construction
`issue. I will note that patent owner advanced a claim
`construction issue for this no-switching criterion in its
`proffer. Your Honors reviewed that and frankly rejected it as
`an attempt to import limitations into the claim. Since
`institution and in the entirety of this trial there has been
`no discussion of claim construction. That issue has been
`abandoned by the patent owner.
` Turning to 6, there's a little bit of a road map.
`We'll go quickly to 7. I want to speak to the combination of
`Golan with Martin. These are ones that you are of course
`familiar with. We have -- yes, Your Honor? Do you have --
` JUDGE MOORE: Something small flew in front of my
`face.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`
`11
`
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`
`IPR2020-00861 (Patent 10,230,898 B2)
`IPR2020-00862 (Patent 10,356,332 B2)1
` MR. O'BRIEN: It happens. So speaking to the
`combination of Golan and Martin, we have established that they
`are analogous art. We have established that both in terms of
`their field of endeavor. The first four rows of this table
`tend to correspond to field of endeavor inquiries. I'll get
`back to that in a second. And we've also established in terms
`of the other portion of that board inquiry which is reasonably
`pertinent to the problem faced by the inventors. That being
`the problem of reducing that jump effect, that me jumping to
`the side when we switch effect that is inherent in parallax
`image systems.
` So on the analogous art issue, the proper inquiry
`of the field of invention is really imaging systems. We have
`in this proceeding and other proceedings made that very clear
`that under KSR guidance as the broad field the patent owner
`has said that's way too broad. And without conceding that
`point, because that is the right analysis, we have made the
`further point that even if one took a much narrower view of
`imaging systems, video output images, two imaging sections,
`parallax images, so narrowing it really to all those
`characteristic aspects, these are in the same field of
`endeavor. And as I said, the proper inquiry is imaging
`systems and however you cut it they are in the same field of
`endeavor. It is also undisputed that the 898 patent itself
`and the 332 patent itself are in that field of endeavor.
` So turning -- I do want to comment on two other
`things that relate to the patent owner's treatment of Martin
`12
`
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`
`IPR2020-00861 (Patent 10,230,898 B2)
`IPR2020-00862 (Patent 10,356,332 B2)1
`and this seems as good a time as any. The patent owner has
`raised for the first time in its slides, Your Honor, two -- at
`least two concepts which were never of record or never argued
`in this proceeding. One is the assertion and you'll see it in
`their slides when they get to them on 12, 13, and several
`others down from there that Martin is only about post-
`processing captured still images. Well, that's wrong. Martin
`is very clearly about both acquisition of video sequences of
`images, sequences of frames, and it's about outputting video
`sequences of images.
` So that's clearly wrong when they try to place that
`in your minds. And I'll just reiterate, that point has never
`been made in this record. They tried to argue it in previous
`proceedings. They argue it for the first time in these
`slides. You'll note that all those slides have no citations
`to the record. There's a reason they don't have citations to
`the record. It isn't in this record. And it's provably
`false.
` Second, patent owner argues that Martin's critical
`alignment only preserves parallax. Judge Moore in a prior
`proceeding and Judge Ullagaddi also in a prior proceeding, you
`asked questions of patent owner's counsel on this point.
`Absolutely Martin eliminates parallax in the region of
`interest. So their rejoinder and their constant rejoinder
`that Martin's critical alignment only preserves parallax,
`again, provably false, and two, not made in this record.
`They're arguing it here in their slides and I presume that
`13
`
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`
`IPR2020-00861 (Patent 10,230,898 B2)
`IPR2020-00862 (Patent 10,356,332 B2)1
`you'll hear from Mr. Fenster when he stands up on that. But
`not in the record, provably false, and there's a reason they
`don't have citations to the record on that.
` So let me turn now to slide 8. Slide 8 lays out
`the KSR motivation to combine case. This is really quite
`straightforward. We do it primarily to provide you with a
`road map into the petition, into the reply for the need, the
`benefit, and the predictable results of the combination
`inquiry that we're required to make under KSR. So that's
`there for you.
` Turning to slide 9. Let me talk briefly to Golan.
`Golan will be familiar to Your Honors. Judge Dougal, perhaps
`this will be the first time you're seeing Golan, and I'll
`spend a little time on it for everyone's benefit, but Golan
`teaches switching again between a wide captured image and a
`tele image in the course of this zoom in, switch, and continue
`to zoom process. So it is within -- it's in the zoom-in
`operation. The switch is performed at a switch point. As I
`described before, it is a switch point at or after which the
`digitally zoomed into portion of the wide is fully subsumed
`within the tele. At or after that point we would be eligible
`to switch because the entire field of view is represented in
`the tele image and we can go on from there.
` As you'll see in the illustration, rectangles for
`wide and tele image. The open rectangle is meant to be the
`wide. The smaller hashed rectangle is meant to be the tele.
`But you'll see this progression. We start from a wide zoom.
`14
`
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`
`IPR2020-00861 (Patent 10,230,898 B2)
`IPR2020-00862 (Patent 10,356,332 B2)1
`We continue to zoom in digitally. At some point, and that is
`at the zoom switch factor, we switch to the tele image. At
`that point in Golan the image is transformed to match the
`other one. And in Golan, we can get to in a moment, that
`transformation is based on the fixed calibration. So the
`previously calculated transformation coefficients that
`transform from the coordinate system of the tele lens and
`imaging sensor, one parallax image, to the other. So that
`transformation occurs. And from there we continue to zoom in.
`Straightforward, that's Golan.
` Turning to slide 10. This jump problem, it is
`solved by the transformation. This -- you know, I keep doing
`that stupid jump across the screen but that's what would occur
`if I jumped from one parallax image captured by one sensor and
`one lens to the other. There's that shift. And that's
`parallax. That jump effect that would occur if I simply
`switched without transforming is well known in the art. It's
`fundamental to systems that have two sensors such that there
`is a parallax. There's a geometry issue that results in
`things closer to the camera, closer to the imaging sensors and
`lenses shifting more, and things further and further away
`shifting less. And at infinity it's essentially no shift.
` But that depth-based amount of shift is parallax
`and the amount of shift is either the parallax shift or the
`parallax effect. Parallax always exists. A shift is
`presented in images and in the teachings that we'll get to is
`eliminated by either the use of the calibration to drive a
`15
`
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`
`IPR2020-00861 (Patent 10,230,898 B2)
`IPR2020-00862 (Patent 10,356,332 B2)1
`transformation in Golan or a proof technique in Martin using
`critical alignment which is an image registration technique to
`do the same thing particularly in the ROI. So point on this
`slide is the jump problems due to parallax images were well
`known regardless of whether it's a zoom system or not a zoom
`system. Your record shows that. Ahiska, Scarff. We pointed
`those out to you here on the slide.
` Turning to slide 11. I'm going to speak to Martin
`I guess for the first time now. Again, this is a reference
`that should be very familiar, maybe too familiar to two of
`you. Martin teaches addressing this very same jump problem
`with a technique that it calls critical alignment. Critical
`alignment effectuates or performs that transformation by
`performing image registration and shifting according to the
`actual images captured.
` So in that sense it's a little different than Golan
`in which a previously captured test target image was using for
`calibration to set those things up. And if the test target
`was a good representation of say me, today's subject of the
`video then it would work pretty well. And if the camera
`hadn't been dropped and we weren't subject to some wild
`temperature swings so that parts tended to move a little bit,
`that would be a pretty good calibration to perform the
`transformation.
` Image registration, the technique that's described
`in Martin and called critical alignment, does have some
`advantages however because it is using the images captured
`16
`
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`
`IPR2020-00861 (Patent 10,230,898 B2)
`IPR2020-00862 (Patent 10,356,332 B2)1
`right now today of me at, you know, 16.7 feet from the camera
`as opposed to more or less 12 where the test target was shot.
`So it has all of those advantages. And Martin teaches, and
`you'll see the illustration. You know, again, two of you have
`seen this many, many times before the illustration in the
`figure 3A, 3B, 3C, 3D sequence.
` 3A is image captured from one sensor. 3B is the
`image captured from the other imaging system and sensor that
`has this shift. So it has this parallax shift. You know, the
`features or the objects and in particular the region of
`interest image which is defined here in Martin as the square,
`not the circle or the triangle. But that collection and
`certainly the ROI is shifted. It's way over on the right side
`whereas it's over down the lower left on the preceding image
`from the other sensor.
` What Martin's critical alignment does is based on
`the image registration it determines the amount of shift and
`you'll see in figure 3C it's a move to the left, I think that
`works for you, and moved down to align with in particular the
`region of interest marker, the square in 3A. And once it's
`done that you'll notice that the resultant position matched
`image in 3D lines up precisely with 3A.
` Now Martin's teaching is that that elimination of
`the parallax shift is performed for the region of interest.
`If it turns out that there are other aspects of the scene that
`are much further back or much closer in the field of view,
`they will not be shifted correctly or sufficiently to entirely
`17
`
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`
`IPR2020-00861 (Patent 10,230,898 B2)
`IPR2020-00862 (Patent 10,356,332 B2)1
`eliminate the parallax shift for those features at shallower
`or deeper depth. But for the one in the ROI, and if I am your
`ROI, you would expect that in the 3A version of the image with
`me and the 3B version image of me that the 3B version would be
`moved, shifted, critically aligned so that the image of me or
`the portion of me as the object in the image would critically
`align as it shows in 3D.
` So that's the concept of critical alignment. We're
`dealing with the stability of me as the region of interest so
`that I'm not in the context of shifting back and forth between
`the image streams that I'm not jumping from side to side or up
`and down or any of those things. That would be unstable.
`That would be distracting to you. It wouldn't be a good
`video. So that's Martin.
` Turning to slide 12. In terms of the combination,
`the reasons to combine, we've road mapped those for you but I
`want to hit this point real hard. It was well known at the
`time that Martin's critical alignment, mainly registration-
`based techniques, were beneficial in systems like Golan if
`calibration, that calculated in advanced, pretty good but not
`perfect (indiscernible), turned out to be insufficient to
`address a jump effect. So it was very well known in the art
`that the technique of Martin, namely critical alignment or
`image registration, had a benefit and could be used to improve
`the performance of systems like Golan.
` That is replete in our record. We cite two major
`references for that and Your Honors have picked up that in
`18
`
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`
`IPR2020-00861 (Patent 10,230,898 B2)
`IPR2020-00862 (Patent 10,356,332 B2)1
`these institutions and other decisions. Ahiska stands
`directly for this proposition. We've provided that
`information for you as well as Border. It's on the slide.
`Both teach the benefit of image registration as an improvement
`when calibration alone is insufficient.
` So our next module I want to go to is Togo. That
`starts on page 13 of the sign post. Again, when I get to 14
`we have again this decomposition again in terms of the
`analogous art inquiry, both in the same field. The first four
`where precedence really should be given to the broad field
`that KSR counsels, the imaging systems field, but even given
`patent owner's protestations, all of these other aspects of
`these systems line up. Imaging systems, they involve the wide
`and the tele sections. They involve the digital zooming. And
`they switch from the wide to the tele during that zoom. So
`Togo has that as does the Golan and Martin combination and of
`course that is what -- that is the field -- the other imaging
`systems where the narrower field I acquiesced, so to speak,
`for the 898. That's also the case for the 332.
` Again, true to form, the problem -- so the
`reasonably pertinent problem that the inventor faced here is
`the effective high-quality imaging of images subject to a
`range of distances. So each of those deals with that problem
`but Togo deals with it in a way that is exactly as the patent
`claims. It introduces a no-switching criterion for them. And
`we discussed that that was an issue.
` So from there I'm going to go to -- on slide 15,
`19
`
`
`
`
`IPR2020-00861 (Patent 10,230,898 B2)
`IPR2020-00862 (Patent 10,356,332 B2)1
`again we do our road map for you in terms of KSR. Again, what
`the need was stated at a high level for effective high-quality
`imaging of objects at range or, more specifically as we sort
`of further detailed in our reply, the question of how to
`achieve high-quality imaging of objects over a range of
`distances and avoid output image degradation when the quality
`of the telephoto image would degrade compared to that of the
`Y. So that's the problem that was well known and I'll get to
`where we know what's well known from. The benefit of the
`solution and the solution here being Togo's use of a no-
`switching criterion and the predictability of the results all
`road mapped for you here on this slide.
` Slide 16 we've already gone through in the context
`of Golan so I'll skip over it. Slide 17, there is no dispute
`that this image quality degradation problem was well known and
`was a well-known issue of telephoto lenses for close subjects.
`Togo addresses this exact problem and it details the problem
`in the Togo reference.
` But we don't need to rely just on Togo. This was
`well known in the art as Dr. Durand explains to you. He
`explains in the quotes that we've provided. He also provides
`you some underlying references to understand how this plays
`out. He provides for you some minimum focus distances for a
`series of Nikon focal lengths for a Nikon zoom lens which
`demonstrate that as the focal length gets longer the minimum
`distance at which that telephoto lens is actually able to
`focus increases.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`
`20
`
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`
`IPR2020-00861 (Patent 10,230,898 B2)
`IPR2020-00862 (Patent 10,356,332 B2)1
` So that exemplifies the problem that Togo
`specifically describes and that people understand in the art
`that often or in general a telephoto lens may not be able to
`focus at as close range as a wide-a