Trials@uspto.gov
`571-272-7822
`
`
`
`
`Paper # 46
`Entered: June 5, 2020
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`____________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`____________
`
`CANON INC., CANON U.S.A., INC., and
`AXIS COMMUNICATIONS AB,
`Petitioner,
`
`v.
`
`AVIGILON FORTRESS CORPORATION
`Patent Owner.
`____________
`
`IPR2019-00311
`IPR2019-00314
`Patent 7,923,923 B2 & C1
`____________
`
`Record of Oral Hearing
`Held: April 8, 2020
`____________
`
`Before GEORGIANNA W. BRADEN, KIMBERLY McGRAW, and
`JESSICA C. KAISER, Administrative Patent Judges.
`
`
`
`
`
`571-272-7822
`
`
`
`
`Paper # 46
`Entered: June 5, 2020
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`____________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`____________
`
`CANON INC., CANON U.S.A., INC., and
`AXIS COMMUNICATIONS AB,
`Petitioner,
`
`v.
`
`AVIGILON FORTRESS CORPORATION
`Patent Owner.
`____________
`
`IPR2019-00311
`IPR2019-00314
`Patent 7,923,923 B2 & C1
`____________
`
`Record of Oral Hearing
`Held: April 8, 2020
`____________
`
`Before GEORGIANNA W. BRADEN, KIMBERLY McGRAW, and
`JESSICA C. KAISER, Administrative Patent Judges.
`
`
`
`
`
`
`IPR2019-00311
`IPR2019-00314
`Patent 7,923,923 B2 & C1
`
`
`APPEARANCES:
`
`ON BEHALF OF PETITIONER AXIS COMMUNICATIONS AB:
`
`
`C. GREGORY GRAMENOPOULOS, ESQUIRE
`Finnegan, Henderson, Farabow, Garrett & Dunner LLP
`901 New York Avenue, NW
`Washington, D.C. 20001-4413
`
`
`ON BEHALF OF PETITIONERS CANON INC. AND CANON U.S.A.
`INC.:
`
`
`RICHARD MARTINELLI, ESQUIRE
`Orrick Herrington & Sutcliffe
`1152 15th Street, NW
`Washington, D.C. 20005
`
`
`
`
`
`ON BEHALF OF THE PATENT OWNER:
`
`
`
`REZA DOKHANCHY, ESQUIRE
`Kirkland & Ellis
`1301 Pennsylvania Avenue, NW
`Washington, D.C. 20004
`
`
`The above-entitled matter came on for hearing on Wednesday,
`
`April 8, 2020, commencing at 1:10 p.m., via Video Teleconference.
`
`
`
`2
`
`
`
`IPR2019-00311
`IPR2019-00314
`Patent 7,923,923 B2 & C1
`
`
`P R O C E E D I N G S
`
`
`JUDGE MCGRAW: This is Judge McGraw and I think we are ready
`
`to begin. So, I would like to start by thanking everyone for your flexibility
`in conducting this remote hearing today. We realize this is a departure from
`our typical practice where we are able to have the parties and at least one
`judge together in the hearing room. Because this is a different type of
`procedure, I'd like to start by clarifying a few items.
`
`First, our primary concern is your right to be heard. If at any time
`during the hearing if you encounter technical or other difficulties that you
`feel fundamentally undermine your ability to adequately represent your
`client, please let us know immediately. For example, by contacting our team
`members who provided you with the connection information.
`
`Second, I'd like to remind everyone that there is no recording of this
`proceeding permitted. Therefore, no one is able to make an audio or a video
`recording. This is the same procedure that we would have in the actual court
`room hearing where nobody is allowed to bring in a recording device into
`the hearing.
`
`Third, when you're not speaking if you could please mute yourself that
`will help keep a clearer record. Fourth, to help the court reporter know who
`is speaking and to prepare an accurate transcript, we ask that everyone
`identify themselves before speaking and the judges, we will also attempt to
`do that each time that we ask a question.
`
`Also, we are very familiar with the record and we have access through
`our computer system to the entire record including the demonstratives.
`
`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
`
`
`
`3
`
`
`
`IPR2019-00311
`IPR2019-00314
`Patent 7,923,923 B2 & C1
`
`When you refer to the demonstratives, however, and you switch slides,
`please be sure to identify the new slide by page number so that we can
`follow along. Also, if you would like the panel to review a particular page
`or a document or exhibit in the record, after you identify that particular page,
`please pause for a few moments as it might take us a few seconds to navigate
`the record to get to that particular page in the record.
`
`Also counsel, because we are doing this remotely, you are not able to
`view the timer that we have in our hearing rooms. So, counsel are urged to
`use some sort of timer to keep track of their argument time. Later, I will get
`each party's rebuttal reserval time and we will make every effort to let you
`know when you are going into rebuttal and to give you a few minutes
`warning before all of your time is up. But please remember it's counsel's
`responsibility to keep track of time.
`
`Finally, please be aware that additional co-counsel, members of the
`public, including client representatives may be listening in to this oral
`hearing through an audio connection. So, please refrain from disclosing any
`confidential information. Does anyone have any questions about the issues
`that I just reviewed? Petitioner Mr. Martinelli?
`
`MR. MARTINELLI: No, Your Honor, that was very clear.
`
`JUDGE MCGRAW: Thank you and Mr., could you please pronounce
`your last name again?
`
`MR. DOKHANCHY: Sure, Your Honor, it's Dokhanchy.
`
`JUDGE MCGRAW: Thank you, Mr. Dokhanchy.
`
`MR. DOKHANCHY: Thank you, Your Honor and no questions from
`us.
`
`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
`
`
`
`4
`
`
`
`IPR2019-00311
`IPR2019-00314
`Patent 7,923,923 B2 & C1
`
`JUDGE MCGRAW: Thank you. As set forth in our oral hearing
`
`order and this is Judge McGraw speaking, each side has 75 minutes total to
`argue. Although the 311 and 314 [dropped audio 3:46] those cases will be
`argued together as they both involve the same patents and there is a
`substantial overlap of issues. A single transcript will be created an entered
`into each proceeding.
`Petitioner who has the burden of proof to show the unpatentability of
`the challenged claims by a preponderance of the evidence will begin by
`presenting its case. Patent Owner will then respond. Thereafter, Petitioner
`may use any time it has reserved for rebuttal to respond to Patent Owner's
`argument. Petitioner's rebuttal time may not be more than half of the
`allotted time. Thereafter, Patent Owner may present a brief surrebuttal.
`With that, I invite Petitioner to begin and how much time would you like to
`reserve for rebuttal?
`
`MR. MARTINELLI: Thank you, Your Honor. I'd like to reserve 35
`minutes.
`
`JUDGE MCGRAW: 35 minutes which means you have 40 minutes
`for your opening presentation.
`
`MR. MARTINELLI: All right, well I'll begin then.
`
`JUDGE MCGRAW: Thank you very much.
`
`MR. MARTINELLI: I thank all of you and all the staff for attending
`in these strange times that we're in and I appreciate everybody making
`themselves available. I'm Richard Martinelli. I will be representing Canon,
`Petitioner Canon in this hearing.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`5
`
`
`
`IPR2019-00311
`IPR2019-00314
`Patent 7,923,923 B2 & C1
`
`I'd like to start with just a brief discussion of the 923 patent to orient
`
`ourselves before we jump into the issues. If you can turn to slide 4 of my
`demonstratives there's just some information giving a basic overview of the
`disclosure of the 923 patent. The 923 patent discloses -- lost the video there
`for a second, sorry everybody. The 923 patent discloses the concept of a
`system where video -- it's fundamentally for video retrieval. And the idea
`that you might want to be able to deal with video in a semantic sort of
`fashion.
`
`So, what the 923 talks about is having a process whereby video is
`analyzed to detect events, objects, attributes. Primarily the attributes of the
`objects are recorded and then later through a different process, a user could
`search the video attributes to try and find an interesting part of video. And
`the reason that the 923 states that this is important to do this this way is
`because if you wanted to look for a particular piece of video, you would
`have to actually maybe in prior art systems go through and scan the entire
`video.
`
`So, imagine if you're, you know, a security guard and you wanted to
`see if some event occurred, somebody went by some area of the system that
`you're monitoring or the site that you're monitoring. You would normally in
`an old school system have to go and just like manually review all of that
`video looking for whatever you're looking for. And the 923 says, wouldn’t
`it be better if we can define some attributes, dictate an event and then go and
`search just for that event.
`
`With that basic parameter out of the way, as was recognized in the
`institution decision, the prior art that we're dealing with here discloses
`
`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
`
`
`
`6
`
`
`
`IPR2019-00311
`IPR2019-00314
`Patent 7,923,923 B2 & C1
`
`exactly that concept. Both the Kellogg and Dimitrova references are
`directed to the idea that you want to look at the semantics of the video, you
`want to record semantic data about the video. You want to record attributes
`that give a user the freedom to then use that semantic data to go in and look
`for any particular piece of video that they might be interested in.
`
`And the institution decision found all of the limitations of the claim
`were reasonably shown at that stage and I think maybe we should spend
`most of our time here talking about what has the Patent Owner identified as
`its distinctions. And I think there is two big issues that I'd like to discuss.
`
`One is the fact that a number of these issues have already been
`decided in a prior IPR between the parties dealing with a related patent to
`the 923 patent. And as we'll discuss in more detail, the Board doesn't need
`to revisit and redecide issues that have already been before it and that have
`already been decided. The doctrine of collateral estoppel is available to
`meet the Board's burden of having to readdress issues that have already been
`addressed.
`
`Then even if we go to the substance of the particular issues and if we
`go to the substance of the issues that aren't subject to collateral estoppel, I
`think what we'll see is the distinctions the Patent Owner is making to try and
`overcome the prior art are all at a level of detail that's inconsistent with the
`specification of the 923 patent. I think what we have here is we have prior
`art references that are very detailed on how to actually implement the
`systems that they disclose. And a patent that has high level description and a
`Patent Owner that's criticizing the prior art (inaudible) details.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`7
`
`
`
`IPR2019-00311
`IPR2019-00314
`Patent 7,923,923 B2 & C1
`
`I think there's a few places that we see this. One is with some of the
`
`constructions the Patent Owner is asking for that are sort of a distortion of
`the (inaudible) and when you look at many of them, they're not supported by
`the specification. Similarly, you know there's features of the prior art that
`are criticized by the Patent Owner but then the patent operates in exactly the
`same way. I think a good example there is trajectory attributes. The Patent
`Owner repeatedly criticizes the prior art for recording trajectory attributes.
`But the patent discloses exactly the same thing and tells you to record
`trajectory attributes.
`
`Then finally, there's features in the prior art that are described in some
`detail that frankly is lacking in the patent. The patent is higher level than the
`prior art is but the Patent Owner attempts to use that to criticize the prior art.
`And I think some examples there are the prior art's disclosure of indexing
`and the prior art's disclosure of the particular data structures it uses to store
`its data. So, you know, the Patent Owner tries to find distinction in these
`areas but there's just no support in the spec to distinguish the prior art at that
`granular level that the Patent Owner's attempting.
`
`JUDGE MCGRAW: And counselor are you still on slide 4 or have
`you moved on to slides 5, 6 and 7?
`
`MR. MARTINELLI: I have not moved on to any slides, I was just
`proceeding on.
`
`JUDGE MCGRAW: Very good.
`
`MR. MARTINELLI: And I'm guessing the way things are going, my
`things will not be super linear so I apologize for that. But actually, we can
`turn to slide 8 right now. And I'd like to briefly discuss some of the
`
`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
`
`
`
`8
`
`
`
`IPR2019-00311
`IPR2019-00314
`Patent 7,923,923 B2 & C1
`
`overarching basic concepts for collateral estoppel. Slide 8 shows the four
`main elements needed to show collateral estoppel. But I think the most
`important thing we need to sort of get a handle on is that collateral estoppel
`is about issue preclusion. So, it's tied to specific issues and it's a doctrine
`that says hey, if a prior tribunal addressed an issue, a future tribunal doesn't
`need to go back and revisit that same issue again.
`
`I think some of the criticisms that I've seen from Patent Owner relate
`to trying to turn it into a claim preclusion thing where they're trying to say
`we're attempting to completely tell the Board that they don't have to do
`anything here and they can just rely on the 661 decision. That's not true.
`Our collateral estoppel issues are tied to particular issues present in this case.
`
`So, if we go through the four elements of collateral estoppel, the first
`one is what I just said, right. Collateral estoppel is about issues and it's
`about showing that the issue in the current action is the same as the issue in
`the prior art.
`
`Number two, the issue was actually litigated in the first action. So,
`that was something where the issue was presented and the parties actually
`addressed the issue and I think that's true of all of the things we're asking for
`with respect to collateral estoppel.
`
`Three is similar but there's a nuanced difference here. It's that the
`final judgement in the first action was necessarily required for the
`determination of the identical issue. I don’t think that's a problem we have
`here but just to give you an example so you can get your head around what
`that is. In a typical patent litigation, you might have issues of infringement
`and the validity that are litigated by the parties. But at the end of the case,
`
`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
`
`
`
`9
`
`
`
`IPR2019-00311
`IPR2019-00314
`Patent 7,923,923 B2 & C1
`
`maybe the defendant shows that they're non-infringing. The parties might
`have hotly debated validity but the court never needs to reach that decision.
` So, in that case collateral estoppel wouldn’t apply. Here, the prior IPR
`was dealing with the exact same validity issues that we're dealing with so we
`don't have that problem. And then finally, the fourth element is that the
`parties or the party that the estoppel was going to apply against was fully
`represented. And again, I think Avigilon had adequate counsel and was
`fully represented in the 661 petition.
`
`If we can turn to slide 9. Slide 9 is a timeline that just shows the panel
`when collateral estoppel really attached here and when the different events
`occurred. So, the bottom half of this timeline is the timing associated with
`the 661 IPR. And you'll see that the final written petition for the 661 came
`in in March of 2019 that was after the petition was filed which was filed
`back in the petition here in the 923 patent the prior IPR was filed.
`
`Now, the actual final written decision typically wouldn’t be a -- or
`wouldn’t necessarily be a final determination. Because a Patent Owner
`could appeal and continue to keep those issues alive. And Avigilon did
`actually file a notice of appeal in July of 2019 but on October 2, 2019
`Avigilon withdrew the 661 IPR appeals. That was about a week before it
`provided its Patent Owner responses here in this proceeding. And at the
`point that Avigilon withdrew that appeal, the final determination of the 661
`panel became completely final.
`
`Another interesting while we're just talking briefly about the 661
`patent. I think it's important to know how the 661 patent is related to the
`923 patent. Because sometimes issues between related patents could get
`
`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
`
`
`
`10
`
`
`
`IPR2019-00311
`IPR2019-00314
`Patent 7,923,923 B2 & C1
`
`changed based on the disclosures. So, the 661 patent is, I don’t know how
`many greats it is but it's a great, great, great grandniece or nephew to the 923
`patent with some intervening (inaudible). I think the critical thing to note is
`that the disclosure provided in the 661 patent was more extensive.
`
`So, hypothetically the 661 patent might have more detail that would
`allow somebody to raise a distinction and to distinguish over the prior art.
`But what we have here is the 923 patent has less detail. So, everything that's
`relevant in the 661 patent to deciding that determination that the 661 patent
`was validated, that narrow patent, it's even broader for the 923.
`
`So, if we proceed to slide 10 and we'll talk about some of these in
`more detail as we get to the different issues. This is just a short sort of
`outline of some of the things that were decided in the 661, heavily litigated
`subject to a fully reasoned opinion. That we think Avigilon is now
`continuing to pursue and that the Board really shouldn't be burdened to
`revisit.
`
`The first issue is the construction of user rule and whether a user rule
`requires a response or not. That exact issue was addressed in full in the 661
`IPR and admittedly here they call it a new user rule. But there's no argument
`that the addition of the word new fundamentally changes what the user rule
`is.
`The second issue which is a big one that we'll talk about is this
`
`independence concept. And again, that goes back to the overview we had of
`what the concept is here. The concept for the 923 is detecting attributes and
`then later defining an event in the context of those attributes. Now that was
`something that again was front and center in the 661, that exact same issue
`
`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
`
`
`
`11
`
`
`
`IPR2019-00311
`IPR2019-00314
`Patent 7,923,923 B2 & C1
`
`and the prior panel found that the prior art teaches that limitation and
`construed the claim as necessary to find that showing.
`
`Next getting into the actual prior art itself there was the ruling that
`both that Kellogg teaches the independence elements. There was a ruling
`that the prior art references that we're combining here, Kellogg and Brill
`would be combined by a person of ordinary skill in the art. And then the
`same for Dimitrova with its independence teaching and Dimitrova with its
`combination.
`
`So, I think that's it for the overview of the collateral estoppel. I was
`going to get into the issue of attributes and events next. Of course, if there's
`anything that the panel wants to hear specifically just let me know.
`
`JUDGE MCGRAW: We are familiar with the record so feel free to
`move along. Thank you.
`
`MR. MARTINELLI: Absolutely. So, with respect to the distinction
`between attributes and events, if you look at slide 12, Patent Owner, and I'm
`doing this first in the context of Kellogg. A lot of the issues are the same
`and we can talk about them in the context of Dimitrova but we'll start with
`Kellogg. When you go through the Patent Owner's responses, I believe
`there's a number of different places where they fundamentally criticize the
`prior art because they say it's detecting events and not attributes. And they
`make that argument for a number of different limitations and so slide 12 just
`sort of collects the different limitations where we saw the Patent Owner
`making the same fundamental argument.
`
`When we get to -- if we can move onto slide 13. Slide 13 lays out just
`the basic intro to Kellogg and what Kellogg is doing. So, Kellogg was the
`
`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
`
`
`
`12
`
`
`
`IPR2019-00311
`IPR2019-00314
`Patent 7,923,923 B2 & C1
`
`thesis of an MIT student and what Kellogg was trying to accomplish was to
`provide a really robust way to store data about video and then to make that
`data about the video retrievable later. And a quote here that I'd like to point
`out, visual memory could act as an interface between inputs and
`applications. So, inputs are the video that you want to analyze and
`applications in a computer system, those are the, you know, actual semantic
`part that wants to look at the video and do some work with it.
`
`And they give an example, for example, computer vision algorithms
`for a security system could analyze data provided by various cameras and
`store information in the visual memory. Applications could then retrieve the
`data to track objects, watch for suspicious events and respond to user
`queries. The visual memory would coordinate this information and
`eliminate the need for full connectivity between the inputs and the outputs.
`
`Again, that's a hallmark of independence here, right? Like they're
`eliminating and they're putting an abstraction layer between the part of the
`system that does the video and the part of the system that actually tries to,
`you know, get the results it wants out of the video.
`
`If we proceed to slide 14, I have here an example of one of the events
`that Kellogg discloses could be searched for. And so, this is part of the
`queryng and the syntax that Kellogg provides. And you'll see this is an ad
`hoc thing. Cutting through some of this code, essentially what it says is you
`have a variable person p, you have a timeline range and then you have a
`second variable person q. And then you do the heart of the work which is
`look for situations where person q comes within three units of person p.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`13
`
`
`
`IPR2019-00311
`IPR2019-00314
`Patent 7,923,923 B2 & C1
`
`So, that is something that a user can do in Kellogg. It's completely ad
`
`hoc and independent from the data that was collected. For example, if I
`came along and wanted to, you know, look for a closer connection between
`the centroid I could write this query and put in a smaller number. If I later
`wanted to search for an intersection of three people, I can frame a query to
`add a third person and put in a different dimension. This is something that's
`completely unknown at the point of the video detection part of Kellogg.
`
`So, if we proceed on to slide 15, slide 15 through 17, I don't know that
`I need to belabor each of them. But they basically collect a variety of
`statements from the 661 petition showing that this issue was directly
`addressed. And the prior panel found that Kellogg practices the independent
`that was described in the patent.
`
`If we go to slide 18, we have the claim construction issue around
`independence. And this is (dropped audio).
`
`JUDGE MCGRAW: I believe that Patent Owner, this is Judge
`McGraw, I believe that Patent Owner has an argument about an issue that
`was not present in the 661 IPR proceedings. And that is Kellogg does not
`teach the “only” limitation as the term “only” was not present in the 661
`claims. And Patent Owner when it's their turn to argue can correct me if I'm
`wrong.
` But I understand their argument to be that when Kellogg performs the
`searches that you're identifying here, such as the approach query, that it is
`not performed on the bucket list index. And because of that the Kellogg
`visual system does not satisfy the “only” limitation that's present in the
`claims at issue in this proceeding.
`
`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
`
`
`
`14
`
`
`
`IPR2019-00311
`IPR2019-00314
`Patent 7,923,923 B2 & C1
`
`MR. MARTINELLI: Okay yes, Judge McGraw, I think that is an
`
`issue and there's a couple points there. One I think is a claim construction
`issue that we'll have to address and then another is what are the fundamental
`teachings of Kellogg. And I'll do the Kellogg teachings first if we could go
`to slide 34 in my presentation.
`
`So, what we rely on to show that Kellogg is searching only the
`attributes is this bucket index that's disclosed in Kellogg. The bucket index
`is basically a quick and dirty plain way of storing the attributes. And the
`way Kellogg describes it is that you'll collect the attributes and you'll just --
`it's like you're throwing them a bucket. You'll just put them in a list that you
`can go back and search later.
`
`Now that raw set of attributes can be searched. If I understand Patent
`Owner's argument with respect to this, they're somehow saying it's in the
`wrong section of Kellogg or it wasn't, you know, the system of Kellogg.
`Well, Kellogg has a bunch of embodiments and the bucket index is one of
`the embodiments. And, in fact, Kellogg actually built the system using the
`bucket index and used it to compare to other ways of organizing the data that
`it taught.
`
`So, if you look on slide 34, there's a nice signal that they have where
`they compare the performance of the bucket index to to the quadtree and
`grid indexes. Now, you know, Kellogg was trying to show some fancy
`technology with the quadtree and grid indexes and they did have advantages
`but that doesn't mean that the bucket index wasn't a potential embodiment.
`And, in fact, Kellogg actually teaches some advantages of the bucket index.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`15
`
`
`
`IPR2019-00311
`IPR2019-00314
`Patent 7,923,923 B2 & C1
`
`So, if we look at page 83 of Kellogg and I'll just read it, I don’t know
`
`if you have to actually pull it up. The bucket index simply maintains a list of
`all the objects stored in the visual memory. Since there are no overheads for
`storage of complex index structures, a bucket index can achieve the highest
`update performance.
`
`So, if you had an embodiment and you were a person of skill in the art
`who wanted to build something where update performance was important,
`Kellogg would tell you, use a bucket index. So, I don't see anything here
`that suggest that the bucket index is not a valid teaching and not a valid
`embodiment. And again, tying it to the approach query, this would apply to
`all the queries. This is a situation where the teachings of Kellogg are
`optional elements that would make sense to be added into any of the
`different embodiments. So, I don't think there's an issue --
`
`JUDGE KAISER: So, Mr. Martinelli, this is Judge Kaiser. I just
`want to pause for a quick second and I apologize for interrupting you. I
`don't see anybody's video now so I just want to make sure that you all can
`hear me and that we still have everybody.
`
`MR. MARTINELLI: So, I can hear you, Judge Kaiser.
`
`MR. DOKHANCHY: This is Reza Dokhanchy. I can hear you as
`well, Judge Kaiser.
`
`
`
`JUDGE KAISER: Okay, all right. Well we'll just sort of proceed on
`the basis that at least we still have the audio. But I encourage anyone to
`raise their hand or to call the alternate number if we have this degrade
`further. And maybe I'll just ask since we are experiencing some technical
`
`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
`
`
`
`16
`
`
`
`IPR2019-00311
`IPR2019-00314
`Patent 7,923,923 B2 & C1
`
`difficulty at least on the video. Counsel, you both received an alternative
`call in number if we have technical difficulties, is that right?
`
`MR. MARTINELLI: Yes, Your Honor.
`
`JUDGE KAISER: Okay. So, I do have a question and sorry for the
`side track into the technical issues.
`
`MR. MARTINELLI: No worries.
`
`JUDGE KAISER: I think we're all doing our best. So, I want to
`make sure I understand your argument, Mr. Martinelli, around what you
`referred to as the various or the different embodiments in Kellogg. Because
`you do have an anticipation ground based on Kellogg. And so, I'm trying to
`understand if your position is that it's okay for you to rely on the different
`embodiments in Kellogg as part of that or if what you're saying is essentially
`there are a bunch of different options in Kellogg that apply sort of regardless
`of the embodiment so sort of a mix and match is okay.
`
`MR. MARTINELLI: Right. So, I think the position is that Kellogg
`teaches a person of skill in the art in these embodiments to use each of these
`things and when to use them. So, if you look at some of the cases where
`this, you know, issue of anticipation comes in, I think it's a very different
`situation than what we have in Kellogg. I think what you have in those
`cases is situations where there's a holistic part. You know, I think one of the
`cases dealt with protocols. And you've got two independent alternatives, say
`two protocols and the district court in that case erred because it took
`subsections of one protocol and subsections of another protocol and mixed
`and matched those together.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`17
`
`
`
`IPR2019-00311
`IPR2019-00314
`Patent 7,923,923 B2 & C1
`
`So, you know, if you were imagining say somebody had a patent
`
`where they disclosed you could use WiFi or Bluetooth and then gave some
`detail about elements of WiFi and elements of Bluetooth. I think it would be
`improper to say I'm going to make an anticipation where I create this new
`hybrid protocol that's neither WiFi nor Bluetooth but some amalgam of the
`two. That's not what we have here.
`
`What we have here is a situation where Kellogg says you can use a
`bucket index, you can use a quadtree, you can use any other index you want
`and all of them are good and useable for their case. So, I'm not taking, you
`know, things that are unique to the bucket index and applying them across to
`something that Kellogg distinguishes. So, the searching that you do on the
`index or on the set of attributes is unrelated to the way you organize those
`attributes.
`
`And again, I think this is another situation where the Patent Owner
`has sort of gotten away from what the 923 patent actually teaches. So,
`there's no basis in the 923 patent to support defined distinctions over how
`you store data. There's actually precious little to support Patent Owner's
`only construction that they're advocating here. And I'm sure the Patent
`Owner will spend some time on this so I don’t want to make their arguments
`for them.
`
`But the only limitation I think they're arguing should be given to plain
`and ordinary meaning and I think the prior art shows “only” under the plain
`and ordinary meaning. The Patent Owner then, despite saying it's the plain
`and ordinary meaning, adds in a bunch of concepts about abstractions that I
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`18
`
`
`
`IPR2019-00311
`IPR2019-00314
`Patent 7,923,923 B2 & C1
`
`find no support for in the specification at all. And I've got a couple slides on
`this.
`So, slide 30 and 31 in my presentation has every reference to the term
`
`abstraction in the 923 patent with the exception of the word abstract on the
`front page of the patent. There's nothing in here that distinguishes
`abstractions and attributes in the way that Patent Owner is attempting to use
`them to try and distinguish the prior art. There's nothing in here about the
`type of indexing you need to use, you know, distinguishing one type of
`indexing from another. There's nothing in here about how you store the
`data, it's just simply not in the specification and not supportable.
`
`I could read a couple of these. So, on slide 30, we've got the
`statement, in block 31, one or more object types of interest are identified in
`terms of video primitives or abstractions thereof. Okay, fair enough.
`Examples of one or more objects include an object, a person, a red object,
`two objects, two persons, a vehicle. So, nothing in there even tells you how
`you distinguish between the video primitives and the abstraction must less,
`you know, the extra gloss that Patent Owner is trying to put on abstractions
`in its only construction.
`
`Then in the next paragraph in block 35, one or more discriminators
`are identified by describing interactions between video primitives or their
`abstractions, spatial areas of interest and temporal attributes of interest.
`Again, nothing in there to say indexes are bad or, you know, what an
`abstraction is or how you distinguish between abstractions or not
`abstractions.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`19
`
`
`
`IPR2019-00311
`IPR20
Accessing this document will incur an additional charge of $.
After purchase, you can access this document again without charge.
Accept $ Charge
Still Working On It
This document is taking longer than usual to download. This can happen if we need to contact the court directly to obtain the document and their servers are running slowly.
Give it another minute or two to complete, and then try the refresh button.
A few More Minutes ... Still Working
It can take up to 5 minutes for us to download a document if the court servers are running slowly.
Thank you for your continued patience.
This document could not be displayed.
We could not find this document within its docket. Please go back to the docket page and check the link. If that does not work, go back to the docket and refresh it to pull the newest information.
Your account does not support viewing this document.
You need a Paid Account to view this document. Click here to change your account type.
Your account does not support viewing this document.
Set your membership
status to view this document.
With a Docket Alarm membership, you'll
get a whole lot more, including:
- Up-to-date information for this case.
- Email alerts whenever there is an update.
- Full text search for other cases.
- Get email alerts whenever a new case matches your search.
One Moment Please
The filing “” is large (MB) and is being downloaded.
Please refresh this page in a few minutes to see if the filing has been downloaded. The filing will also be emailed to you when the download completes.
Your document is on its way!
If you do not receive the document in five minutes, contact support at support@docketalarm.com.
Sealed Document
We are unable to display this document, it may be under a court ordered seal.
If you have proper credentials to access the file, you may proceed directly to the court's system using your government issued username and password.
Access Government Site
