1 UNITED STATES PATENT AND TRADEMARK OFFICE
`
`2 BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`3 - - - - - - - - - - - - - - - - -x
`
`4 CANON INC., CANON U.S.A., INC., :
`
`5 and AXIS COMMUNICATIONS AB, :
`
`6 Petitioner, : CASE NO'S.
`
`7 v. : IPR2019-00311
`
`8 AVIGILON FORTRESS CORPORATION, : IPR2019-00314
`
`9 Patent owner. : PATENT NO.
` 7,932,923
`10 - - - - - - - - - - - - - - - - -x
`
`11
`
`12
`
`13 Videotaped Deposition of JOHN GRINDON, D.Sc
`
`14 Washington, D.C.
`
`15 Wednesday, October 2, 2019
`
`16 9:36 a.m.
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23 Reported by:
`
`24 Cassandra E. Ellis, RPR
`
`25 Job No.: 10060621
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 1 of 187
`
`
`
`2 BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`3 - - - - - - - - - - - - - - - - -x
`
`4 CANON INC., CANON U.S.A., INC., :
`
`5 and AXIS COMMUNICATIONS AB, :
`
`6 Petitioner, : CASE NO'S.
`
`7 v. : IPR2019-00311
`
`8 AVIGILON FORTRESS CORPORATION, : IPR2019-00314
`
`9 Patent owner. : PATENT NO.
` 7,932,923
`10 - - - - - - - - - - - - - - - - -x
`
`11
`
`12
`
`13 Videotaped Deposition of JOHN GRINDON, D.Sc
`
`14 Washington, D.C.
`
`15 Wednesday, October 2, 2019
`
`16 9:36 a.m.
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23 Reported by:
`
`24 Cassandra E. Ellis, RPR
`
`25 Job No.: 10060621
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 1 of 187
`
`
`
`1 Videotaped Deposition of JOHN GRINDON, D.Sc, held at the
`
`2 offices of Kirkland & Ellis, LLP, 1301 Pennsylvania
`
`3 Avenue, Northwest, Washington, D.C. 20004, pursuant
`
`4 to agreement, before Cassandra E. Ellis, Certified
`
`5 Court Reporter - Virginia, Certified Court Reporter -
`
`6 Washington, Certified Shorthand Reporter - Hawaii,
`
`7 Registered Professional Reporter, Certified Livenote
`
`8 Reporter, Realtime Systems Administrator, Registered
`
`9 Professional Reporter, and Notary Public of The
`
`10 District of Columbia.
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 2 of 187
`
`
`
`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
`
` A P P E A R A N C E S
`
`ON BEHALF OF PETITIONER CANON:
`
`RICHARD MARTINELLI, ESQUIRE
`
`ORRICK, HERRINGTON & SUTCLIFFE, LLP
`
`51 West 52nd Street
`
`New York, New York 10019
`
`(212) 506-3702
`
` Rmartinelli@orrick.com
`
`ON BEHALF OF PETITIONER AXIS:
`
`GUANG-YU ZHU, ESQUIRE
`
`FINNEGAN, HENDERSON, FARABOW,
`
`GARRETT & DUNNER, LLP
`
`901 New York Avenue, Northwest
`
`Washington, D.C. 20001
`
`(202) 408-4000
`
` Guang-yu.zhu@finnegan.com
`
`ON BEHALF OF PATENT OWNER:
`
`REZA DOKHANCHY, ESQUIRE
`
`KIRKLAND & ELLIS LLP
`
`555 California Street
`
`San Francisco, California 94104
`
`(415) 439-1469
`
`Reza.dokhanchy@kirkland.com
`
`ALSO PRESENT: Robyn A. Ellis, Legal Videographer
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 3 of 187
`
`
`
` C O N T E N T S
`
`EXAMINATION OF JOHN GRINDON, D.Sc
`
`By Mr. Dokhanchy
`
` PAGE
`
`6
`
`E X H I B I T S
`
` (Attached to the Transcript)
`
`JOHN GRINDON, D.Sc Deposition Exhibit
`
` PAGE
`
`Exhibit 1001 US Patent 7,932,923 B2
`
`Exhibit 1003 Visual Memory by Christopher
`
`46
`
`13
`
`James Kellogg
`
`Exhibit 1006 Transactions Information Systems
`
`129
`
`Special Issue on Video Information Retrieval
`
`Exhibit 1016 Video Surveillance System
`
` 102
`
`Employing Video Primitives Amendment and
`
`Reply
`
`Exhibit 1022 Object-Oriented Conceptual
`
`122
`
`Modeling of Video Data
`
`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
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 4 of 187
`
`
`
`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
`
`P R O C E E D I N G S
`
` THE VIDEOGRAPHER: Good
`
`morning. This is the beginning of
`
`disk number one in the deposition
`
`of Dr. John Grindon, taken in the
`
`matter of Canon, Inc., Canon USA,
`
`and Axis Communications AB,
`
`petitioner, versus Avigilon
`
`Fortress Corporation, patent
`
`owner, Case Number IPR2019-00311
`
`and IPR2019-00314, held in the
`
`United States Patent and Trademark
`
`Office before the Patent Trial and
`
`Appeal Board.
`
`Today's date is October 2nd,
`
`2019, and the time on the monitor
`
`is 9:36 a.m. My name is Robyn
`
`Ellis, and I am the legal
`
`videographer, the court reporter
`
`is Cassandra Ellis. We are
`
`representing Aptus Court
`
`Reporting.
`
` Counsel appearances will be
`
`noted on the stenographic record.
`
`Will the court reporter
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 5 of 187
`
`
`
`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
`
`please swear in the witness then
`
`you may proceed.
`
` JOHN GRINDON, D.Sc
`
`having been sworn, testified as follows:
`
` EXAMINATION
`
`BY MR. DOKHANCHY:
`
`Q
`
`A
`
`Q
`
`Good morning, Dr. Grindon.
`
`Good morning.
`
`You've been deposed several
`
`times before; is that right?
`
`A Yes.
`
` Q Can you approximate how many
`
`times?
`
`A I don't have an exact count. I
`
`would say 15 to 20, something like that.
`
` Q Okay. So I'm sure you're
`
`familiar with the basic mechanics of a
`
`deposition, but if anything is unclear
`
`about the procedure or about any of my
`
`questions, just feel free to let me know,
`
`okay?
`
`A Will do.
`
` Q All right. I'll just ask that
`
`we not talk over each other, so that the
`
`court reporter can get everything down,
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 6 of 187
`
`
`
`1 is that all right?
`
`2 A Sure.
`
`3 Q Is there any reason you cannot
`
`4 give truthful and accurate testimony
`
`5 today?
`
`6 A No.
`
`7 Q Could you please state and
`
`8 spell your full name, for the record?
`
`9 A Yes, John Grindon, J-o-h-n,
`
`10 G-r-i-n-d-o-n.
`
`11 Q Okay. Dr. Grindon, you
`
`12 understand that you're here to testify in
`
`13 connection with US patent number 9 --
`
`14 excuse me -- 7,932,923; is that right?
`
`15 A Yes.
`
`16 Q And if I call that the `923
`
`17 patent, you'll know what I'm talking
`
`18 about?
`
`19 A Yes.
`
`20 Q And you understand that the two
`
`21 IPR proceedings you're here to testify
`
`22 about are IPR2019-311 and -314; correct?
`
`23 A Yes.
`
`24 Q And if I call those the 311 and
`
`25 314 proceedings you'll know what I'm
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 7 of 187
`
`
`
`1 talking about; right?
`
`2 A Yes.
`
`3 Q Okay. When were you retained
`
`4 to work on the 311 and 314 IPRs?
`
`5 A I don't have an exact date.
`
`6 It's been something over a year, I
`
`7 believe.
`
`8 Q Okay. When did you first
`
`9 review the `923 patent?
`
`10 A Again, I don't have an exact
`
`11 date, but it's been something over a
`
`12 year.
`
`13 Q Okay. Had you reviewed the
`
`14 `923 patent before you were retained to
`
`15 work on these two IPR matters?
`
`16 A I -- I may have, I'd have to go
`
`17 back and refresh my memory on that.
`
`18 Q Okay.
`
`19 A I don't recall right now.
`
`20 Q Okay. And I see from your CV
`
`21 you have a substantial engineering
`
`22 background; were any of your jobs
`
`23 specific to surveillance systems?
`
`24 A Specific to surveillance?
`
`25 Q Correct.
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 8 of 187
`
`
`
`1 A I would say yes. Now, when I
`
`2 worked with McDonnell Douglas I was
`
`3 working on the cruise missile program,
`
`4 and part of the work there was to develop
`
`5 a video system to scan the terrain and
`
`6 determine the position of the missile
`
`7 relative to objects that it found on the
`
`8 terrain and so forth.
`
`9 Q Okay. Did that involve
`
`10 recognizing human activity in any way?
`
`11 A That particular one -- could
`
`12 you tell me what you mean by human
`
`13 activity?
`
`14 Q I don't mean anything other
`
`15 than the ordinary usage of the words,
`
`16 "human activity."
`
`17 A Okay. I don't see human
`
`18 activity in the claims, that's why I'm
`
`19 asking what your meaning is of that term.
`
`20 Also, as part of that work, I developed
`
`21 algorithms for determining and
`
`22 recognizing targets, such as ships at
`
`23 sea. So this would involve human
`
`24 activity, the navigation of the ships.
`
`25 I also worked on a missile
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 9 of 187
`
`
`
`1 system, using laser radar, known as
`
`2 LIDAR, L-I-D-A-R, and again, developed
`
`3 algorithms to detect human-constructed
`
`4 objects in the scene.
`
`5 Q And so you agree that the `923
`
`6 patent concerns recognizing objects and
`
`7 then later analyzing those objects; do
`
`8 you agree with that general premise?
`
`9 MR. MARTINELLI: Objection
`
`10 to the form.
`
`11 A The -- the general premise, did
`
`12 you -- can you state that again, please?
`
`13 Q Sure. The `923 patent concerns
`
`14 detecting objects, including attributes
`
`15 of the objects, and then later analyzing
`
`16 those attributes to determine events; do
`
`17 you agree with that?
`
`18 A When you say: "Later
`
`19 analyzing," what do you -- what do you
`
`20 mean by analyzing? What do you include
`
`21 in that?
`
`22 Q I just mean however you
`
`23 understand the term analyzing.
`
`24 A Well, what I would prefer to do
`
`25 is maybe to look at the patents and --
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 10 of 187
`
`
`
`1 and see what the patent says.
`
`2 Q Okay. So without looking at
`
`3 the patent you can't agree or disagree
`
`4 with my statement; is that right?
`
`5 A The -- the patents analyzes
`
`6 video to derive attributes and then,
`
`7 later, there are rules applied to the
`
`8 attributes to detect events. If
`
`9 that's -- if that's what you were saying,
`
`10 then I would agree.
`
`11 Q Okay. Does any of your
`
`12 experience in the engineering field
`
`13 relate to detecting attributes from a
`
`14 video and then applying rules to those
`
`15 attributes to detect events?
`
`16 A Yes.
`
`17 Q Okay. Can you please explain?
`
`18 A This would be some of the
`
`19 research, again, at McDonnell Douglas,
`
`20 among other things. But in developing
`
`21 algorithms and techniques for recognizing
`
`22 objects from video imagery, for the
`
`23 purposes of cruise missile guidance, and
`
`24 then determining, for example, after
`
`25 detecting an object, as I mentioned a
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 11 of 187
`
`
`
`1 moment ago, a shipwreck ignition,
`
`2 automatic target recognition, extracting
`
`3 measurements from the object and going
`
`4 through a classification process to
`
`5 determine whether this was a, for
`
`6 example, a friend or foe ship. So that's
`
`7 one example.
`
`8 Q Okay. In that context, what
`
`9 were the events that would have been
`
`10 determined based on applying rules?
`
`11 MR. MARTINELLI: Objection,
`
`12 form.
`
`13 A The -- the video would be
`
`14 analyzed to determine the attributes,
`
`15 define objects and determine the
`
`16 attributes of the object, and then to --
`
`17 in the -- in the case of the ship
`
`18 example, the appearance of a -- of a
`
`19 particular type of ship in the video
`
`20 would be one event that was determined,
`
`21 the time that it appeared, so the -- the
`
`22 appearance of a particular type of object
`
`23 at a particular time would be one
`
`24 example.
`
`25 Q Okay. One of the references
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 12 of 187
`
`
`
`1 that you analyzed in your declaration
`
`2 that you submitted in connection with the
`
`3 311 and 314 IPRs is called the Kellogg
`
`4 reference; is that right?
`
`5 A Yes.
`
`6 MR. DOKHANCHY: All right.
`
`7 Let me hand you a copy of that.
`
`8 And this is Exhibit 1003, the
`
`9 Kellogg reference.
`
`10 (Previously marked Exhibit
`
`11 No. 1003 was identified for the
`
`12 record.)
`
`13 BY MR. DOKHANCHY:
`
`14 Q And this is the reference that
`
`15 you opined on in your declaration; is
`
`16 that right?
`
`17 A This appears to be the one,
`
`18 yes.
`
`19 Q Okay. And Kellogg uses an
`
`20 object-oriented database; correct?
`
`21 A That's a fair characterization.
`
`22 Q What is your understanding of
`
`23 what an object-oriented database is?
`
`24 A I think what we should do is
`
`25 take a look at, specifically, what
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 13 of 187
`
`
`
`1 Kellogg is describing as his particular
`
`2 database.
`
`3 Q Okay. And we can do that, but
`
`4 first I want to get your understanding.
`
`5 Do you have an understanding of what an
`
`6 object-oriented database is?
`
`7 A Again, I would prefer to look
`
`8 at Kellogg's understanding and
`
`9 specification of what he puts in the
`
`10 database.
`
`11 Q Okay. Can you answer my
`
`12 question: Do you have an understanding
`
`13 of what an object-oriented database is?
`
`14 A An object-oriented database
`
`15 would have to do with the object or the
`
`16 particular -- the particular object, as
`
`17 opposed to the identifier of -- let's
`
`18 see -- I'm having trouble coming up with
`
`19 a concise definition. I don't know if I
`
`20 defined that in my declaration or not.
`
`21 We could take a look. Why don't we do
`
`22 that?
`
`23 Q Okay. So you want to look at
`
`24 your declaration for a definition of
`
`25 object-oriented database; is that what
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 14 of 187
`
`
`
`1 you're doing?
`
`2 A I don't know that I defined it
`
`3 in there, but I'm thinking of how Kellogg
`
`4 is using the -- the visual database.
`
`5 Q Okay. I'm not sure that you
`
`6 provided a definition, but I guess
`
`7 without flipping through your
`
`8 declaration, which is several hundred
`
`9 pages, is there anything you can tell me
`
`10 about how an object-oriented database
`
`11 works?
`
`12 A In a database that is organized
`
`13 by the content or subject of a -- for
`
`14 example, in this case, a video, which
`
`15 then would give you the attributes and
`
`16 other characteristics of the object,
`
`17 would be a characterization.
`
`18 Q Okay. And with respect to
`
`19 those items, the attributes and
`
`20 characteristics of the object are
`
`21 structured around the object itself;
`
`22 right?
`
`23 MR. MARTINELLI: Objection,
`
`24 form.
`
`25 A That sounds like a fair
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 15 of 187
`
`
`
`1 statement.
`
`2 Q And the way that you would
`
`3 access those attributes are to access
`
`4 them through the object data structure;
`
`5 right?
`
`6 A That's essentially right.
`
`7 Q Okay. So in Kellogg there are
`
`8 a number of objects described, and we can
`
`9 feel free to look at them, starting on
`
`10 page 22. There are three main types of
`
`11 objects in Kellogg, that is a spatial
`
`12 object, a temporal object, and a
`
`13 spatiotemporal object; correct?
`
`14 A Yes.
`
`15 Q And, for example, a
`
`16 spatiotemporal object is what you would
`
`17 use to represent a moving object or a
`
`18 moving person; right?
`
`19 A You could use that to describe
`
`20 that, so that would be the joint spatial
`
`21 temporal aspects of an object as opposed
`
`22 to each one independently.
`
`23 Q And so, for example, the
`
`24 spatiotemporal object is what you would
`
`25 use to represent a persing -- person
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 16 of 187
`
`
`
`1 walking down a hall; right?
`
`2 A That -- in general, yes.
`
`3 Q And turning back to the spatial
`
`4 object, the spatial object in Kellogg is
`
`5 defined by a set of points; correct?
`
`6 A Can you -- I don't know that I
`
`7 would just limit it to a set of points.
`
`8 Can you show me where you're looking?
`
`9 Q Sure. If you turn to page 24,
`
`10 if you go to the second to last
`
`11 paragraph, it says: A spatial object is
`
`12 defined by a set of points and a local
`
`13 coordinate system. This information is
`
`14 sufficient to fully represent a spatial
`
`15 object; do you see that?
`
`16 A Yes.
`
`17 Q Okay. And so you agree that in
`
`18 Kellogg a spatial object is defined by a
`
`19 set of points and a local coordinate
`
`20 system; right?
`
`21 A And a local coordinate system.
`
`22 Q And Kellogg says that
`
`23 additional information, such as size,
`
`24 measurements, and other factors, besides
`
`25 the set of points and local coordinate
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 17 of 187
`
`
`
`1 system, is then derivable from that
`
`2 information; right?
`
`3 A Essentially, the last sentence
`
`4 of the paragraph says: Additional
`
`5 information, such as centroid,
`
`6 orientation, and bounding box, is derived
`
`7 from this information.
`
`8 Q And that additional information
`
`9 is not actually stored in connection with
`
`10 the object in Kellogg; right?
`
`11 A I believe that's correct, that
`
`12 that's derivable. But whether I --
`
`13 whether that's then stored anywhere or
`
`14 not, I can't say right now. I'd have to
`
`15 look through Kellogg.
`
`16 Q Okay. But you agree that this
`
`17 section that we're looking at -- and feel
`
`18 free to look through it -- does not
`
`19 describe storing that information, but
`
`20 only that it can be derived from what is
`
`21 stored; correct?
`
`22 MR. MARTINELLI: Objection,
`
`23 form.
`
`24 A This doesn't say it's not
`
`25 stored. I believe, to the extent that
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 18 of 187
`
`
`
`1 these are attributes, Kellogg does
`
`2 disclose storing of attributes. Your
`
`3 question was in this particular section,
`
`4 and I don't see in this particular
`
`5 paragraph that you pointed out that it
`
`6 really goes into storage at all.
`
`7 Q And are you aware of any other
`
`8 part of Kellogg that describes storing
`
`9 those attributes, such as length, surface
`
`10 area, volume, centroid, orientation,
`
`11 bounding box or the like?
`
`12 A I'd have to go through and see
`
`13 where you would talk about storing
`
`14 attributes. I can do that, if you'd
`
`15 like.
`
`16 Q Sure. Why don't you take a
`
`17 look.
`
`18 A On page 50, section 3.6, object
`
`19 storage, through page 52, I believe, it
`
`20 talks about the -- the storage.
`
`21 Q Okay. Can you point me to a
`
`22 passage?
`
`23 A Well, let me look through and
`
`24 pick out a few things, then. So in the
`
`25 preamble to that section it says: The
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 19 of 187
`
`
`
`1 object-oriented database on which the
`
`2 visual memory builds provides basic
`
`3 support for object storage and retrieval.
`
`4 It says that the visual memory assigns
`
`5 each object and object identifier, and a
`
`6 given object can have multiple versions.
`
`7 It says, for example, a
`
`8 security system could track a person
`
`9 walking down a hall and store a new
`
`10 version describing that person's location
`
`11 every tenth of a second. Then it says,
`
`12 by maintaining all of an object's
`
`13 versions the visual memory can answer
`
`14 questions about the object's history. So
`
`15 that would be storing all of these states
`
`16 of the object every certain amount of
`
`17 time, and then using that later, in a
`
`18 query, to determine the history of the
`
`19 object.
`
`20 And the section goes on for a
`
`21 few pages, here, but there's several
`
`22 indications here of storage, then section
`
`23 3.6.2 specifically describes the storage
`
`24 mechanism, before that it talks about
`
`25 objects appearing and disappearing, and
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 20 of 187
`
`
`
`1 that that's appropriately stored, that
`
`2 information.
`
`3 So in this -- this goes into
`
`4 details of -- of the storage concept that
`
`5 Kellogg has, in the second paragraph it
`
`6 talks about in certain cases the database
`
`7 should store one base version of the
`
`8 object and then indicate differences for
`
`9 each new version. That's in the case
`
`10 where the object changes very little from
`
`11 moment to moment.
`
`12 Q Right. So none of that,
`
`13 though, describes storing the attributes
`
`14 that we were just talking about, which
`
`15 are, for example, centroid, orientation,
`
`16 length, width, height, area, that sort of
`
`17 thing; right?
`
`18 A Let me go back, I think I
`
`19 mentioned this a moment ago, but it says:
`
`20 Each object can have multiple versions,
`
`21 for example, a security system could
`
`22 track a person walking. So a person
`
`23 could be an attribute. A person walking,
`
`24 that would be another attribute.
`
`25 And it's storing all of this
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 21 of 187
`
`
`
`1 information, as I just explained a few
`
`2 moments ago. So at least in this section
`
`3 it does indicate that attributes are
`
`4 stored. And we can go through the rest
`
`5 of the Kellogg reference and perhaps find
`
`6 other indications of that.
`
`7 Q Okay. Why don't we turn back
`
`8 to page 24, because I just want to ask
`
`9 about the specific attributes that I was
`
`10 asking about, which, if you go to the end
`
`11 of page 24, and onto 25 it says:
`
`12 Concrete spatial objects can provide
`
`13 additional relevant information; for
`
`14 example, a cube could have functions
`
`15 returning the length of its side, its
`
`16 surface area, and its volume; right?
`
`17 A Yes.
`
`18 Q So as to those variables, what
`
`19 Kellogg teaches is that that information
`
`20 is only obtainable through function calls
`
`21 as defined in the object class; right?
`
`22 MR. MARTINELLI: Object to
`
`23 form.
`
`24 A Can you show me -- can you show
`
`25 me where it says that it can only do
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 22 of 187
`
`
`
`1 that?
`
`2 Q Well, that's -- the part that
`
`3 we just read is where it says it can do
`
`4 that; right?
`
`5 A Can do that, a moment ago you
`
`6 said only.
`
`7 Q Right. And then so I'd asked
`
`8 you to find any other place where there
`
`9 are other ways to obtain that
`
`10 information. And we went to, I believe
`
`11 it was page 50, and none of these
`
`12 attributes were discussed on page 50; do
`
`13 you agree?
`
`14 A On page 50, those were not an
`
`15 exhaustive list. Those were just
`
`16 examples.
`
`17 Q Okay.
`
`18 A And there's more, for example,
`
`19 on page 52. As a historical database,
`
`20 the visual memory keeps track of when
`
`21 events happened -- this is with regard to
`
`22 the time aspect -- it stores with each
`
`23 version of a temporal object information
`
`24 about that version's time. So this is
`
`25 additional discussion about the database
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 23 of 187
`
`
`
`1 storing various attributes of an object.
`
`2 Q Right. So none of the
`
`3 attributes that we were discussing on
`
`4 pages 24 to 25, though, are discussed on
`
`5 page 52; right?
`
`6 A These examples, on page 24 and
`
`7 25, are not exhaustive. The examples on
`
`8 page 52 are not exhaustive, nor page 50
`
`9 and 51, as we discussed earlier.
`
`10 Q Okay. But you're not able to
`
`11 point me to any part of Kellogg that
`
`12 describes obtaining parameters related to
`
`13 length, surface area, volume or other
`
`14 measurement information, other than
`
`15 what's described on pages 24 to 25, using
`
`16 function calls; right?
`
`17 A And I just went through pages
`
`18 50 to 52, which do talk about these
`
`19 things. They talk about object movement,
`
`20 positions of objects, so that this --
`
`21 these are measurements of time and space.
`
`22 Q Okay. But they're not
`
`23 measurements of what we've been talking
`
`24 about on pages 24 to 25; do you agree
`
`25 with that?
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 24 of 187
`
`
`
`1 A I can't agree with that,
`
`2 exactly. For one thing, on page 50, it
`
`3 talked about a person walking. So to
`
`4 determine it's a person, there has to be
`
`5 various information measured from the
`
`6 detected object. And the fact that the
`
`7 person is walking would indicate that
`
`8 there's some measurement of position,
`
`9 which is often done with centroid. So
`
`10 the various low-level parameters that you
`
`11 show on page 24 are elements of the --
`
`12 the higher level of a person or of the
`
`13 activity of a person walking.
`
`14 Q Okay. Will you agree that in
`
`15 order to determine those things on page
`
`16 50 it's not necessary that the items on
`
`17 page 24 to 25 are actually stored, they
`
`18 can just be retrieved using a function;
`
`19 right?
`
`20 A I can't say, right now, that
`
`21 these lower-level attributes, such as
`
`22 centroid, orientation, are -- are not
`
`23 stored. They can be attributes as well
`
`24 as a person walking.
`
`25 So it's not as though we're
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 25 of 187
`
`
`
`1 saying that only this set of attributes
`
`2 can be stored. Kellogg is offering a --
`
`3 a general approach, here. And I think
`
`4 the user can decide whether or not these
`
`5 things will be stored. And I don't think
`
`6 Kellogg is -- is trying to lock the user
`
`7 down into -- into just this list of
`
`8 things to be stored.
`
`9 Q Okay. And again, you agree
`
`10 that those things you pointed out as
`
`11 attributes, such as person and walking,
`
`12 are only accessible by the system through
`
`13 the data structure of the object that
`
`14 they are associated with; correct?
`
`15 A I didn't say that, no.
`
`16 Q Okay. Do you agree with that?
`
`17 A For example, you mentioned
`
`18 person walking.
`
`19 Q Mm-hmm.
`
`20 A So an -- if person is
`
`21 considered an attribute, and walking is
`
`22 considered an attribute of an object,
`
`23 then these could be stored and they would
`
`24 be accessible as -- as attributes.
`
`25 Q Right. My question is: The
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 26 of 187
`
`
`
`1 attributes in Kellogg are not accessible
`
`2 as stand-alone variables. They're only
`
`3 accessible through the data structure of
`
`4 the object they're associated with by
`
`5 virtue of being an object-oriented model;
`
`6 true?
`
`7 MR. MARTINELLI: Objection,
`
`8 form.
`
`9 A The database provides these
`
`10 attributes which are stored. The
`
`11 attributes are directly accessible
`
`12 through the database. So there always
`
`13 has to be some way, in any kind of a
`
`14 database, to access the item that you
`
`15 want. And it's -- he doesn't say it's
`
`16 only accessible through that. In fact,
`
`17 let me look further, we've just been
`
`18 looking at page 24, and then the few
`
`19 pages that I mentioned on data storage,
`
`20 because that was the question you had
`
`21 earlier.
`
`22 But let's look at a couple of
`
`23 things, for one, I want to get back to
`
`24 the Kellogg. Let me see what's in the
`
`25 report on that. Well, there's part of
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 27 of 187
`
`
`
`1 the answer on page 22 of Kellogg, it
`
`2 says: The visual memory design extends
`
`3 the database's storage mechanism. It
`
`4 provides a mechanism for object identity
`
`5 and maintains the history for each
`
`6 object. And then, the last sentence in
`
`7 that paragraph, the design lets
`
`8 applications customize the database
`
`9 storage server based on characteristics
`
`10 of the data they typically store.
`
`11 Q Right. So my question is:
`
`12 When the system wants to access a
`
`13 particular attribute it does so only
`
`14 through the data structure of the object;
`
`15 correct?
`
`16 A And -- no, that's not correct.
`
`17 Q Okay. Please point me to where
`
`18 it discusses doing something other than
`
`19 that.
`
`20 A We'll get to it.
`
`21 Q Dr. Grindon, do you have
`
`22 something specific in mind that you're
`
`23 looking for or are you just looking
`
`24 throughout the Kellogg reference?
`
`25 A I am looking for specific
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 28 of 187
`
`
`
`1 things.
`
`2 Q Okay. What are you looking
`
`3 for?
`
`4 A In the implementation Kellogg
`
`5 shows that he creates a bucket of
`
`6 attributes. I was looking for that
`
`7 particular page. I have it referenced in
`
`8 my declaration.
`
`9 Q Okay. Are you talking about
`
`10 the bucket index?
`
`11 A The bucket index, yeah. I have
`
`12 that section here. I'm looking in my
`
`13 declaration, right now, for specific
`
`14 reference to that.
`
`15 Q I might be able to short
`
`16 circuit this. If you go to page 83 of
`
`17 Kellogg, it references the bucket index.
`
`18 Are you there?
`
`19 A I'm on page 83.
`
`20 Q If you go down to the last
`
`21 paragraph, it says: In addition to the
`
`22 indices described above, each test also
`
`23 includes a bucket index. A bucket index
`
`24 simply maintains a list of all the
`
`25 objects stored in the visual memory; do
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 29 of 187
`
`
`
`1 you see that?
`
`2 A Yes.
`
`3 Q Is that what you're referring
`
`4 to?
`
`5 A That's one of the places, yes.
`
`6 Q Now, the bucket index does not
`
`7 maintain a list of attributes, it
`
`8 maintains a list of objects; correct?
`
`9 Do you understand my question,
`
`10 Dr. Grindon?
`
`11 A Yes, I'm just reading this.
`
`12 Q Do you agree with my question:
`
`13 A bucket index maintains a list of
`
`14 objects, not a list of attributes;
`
`15 correct?
`
`16 A I'm -- I'm looking at this.
`
`17 Q Well, the paragraph we just
`
`18 read, at the bottom of 83, literally
`
`19 says: A bucket index simply maintains a
`
`20 list of all the objects stored in the
`
`21 visual memory; right?
`
`22 Can you answer my question?
`
`23 A Your question was limited to
`
`24 that sentence on page 83, and yes, you
`
`25 read it correctly.
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 30 of 187
`
`
`
`1 Q Okay. You've been looking at
`
`2 this section for awhile, now; is there
`
`3 anything that contradicts that statement?
`
`4 A That's what I was looking for.
`
`5 So on page 75, for example, it
`
`6 says: Temporal indices, section 4.3.3,
`
`7 the prototype temporal indices keep track
`
`8 of the valid times of version -- of
`
`9 object versions. So this index keeps
`
`10 track of the time, which would be an
`
`11 attribute.
`
`12 Q Okay. Can you answer my
`
`13 question, please? When the system wants
`
`14 to access a particular attribute in
`
`15 Kellogg, it does so only through the data
`
`16 structure of the object; correct?
`
`17 A And that appears not to be
`
`18 right. Then, on page 72, under spatial
`
`19 indices, the prototype spatial indices
`
`20 store information about the centroids of
`
`21 objects stored in the visual memory. So
`
`22 these are attributes. In fact, it's one
`
`23 that you had mentioned earlier.
`
`24 Q Where are you looking?
`
`25 A Again, on page 72, section
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 31 of 187
`
`
`
`1 4.3.2, spatial indices, i-n-d-i-c-e-s.
`
`2 Q Mm-hmm. And how does that
`
`3 contradict my question or how does that
`
`4 -- why does that cause you to disagree
`
`5 with my question?
`
`6 A This spatial index stores the
`
`7 information about the centroids of
`
`8 objects.
`
`9 Q Okay.
`
`10 A So it does not say that you
`
`11 have to go through the object. It says
`
`12 that you've got the information about the
`
`13 centroid. And, of course, the attributes
`
`14 are associated with objects.
`
`15 Q And to access the information
`
`16 about the centroid you would then have to
`
`17 go through the index; correct?
`
`18 A When you say: Go through the
`
`19 index, the index would provide that
`
`20 information.
`
`21 Q Right. The centroid of the
`
`22 object does not exist as a stand-alone
`
`23 variable that the database can access
`
`24 without using the index; right?
`
`25 A So I don't see anywhere in the
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 32 of 187
`
`
`
`1 claims, and again, we should put the
`
`2 patent in front of us, where it says that
`
`3 you have to do -- you have to access it
`
`4 the way you said.
`
`5 Q Okay. Well, you may disagree
`
`6 that it's relevant, but my question is
`
`7 just about Kellogg. The centroid of the
`
`8 object does not exist as a standalone
`
`9 variable that the database can access
`
`10 without using the index or the object;
`
`11 right?
`
`12 A Without using the --
`
`13 MR. MARTINELLI: Objection
`
`14 to form.
`
`15 A Without using the index? I
`
`16 don't think there's any requirement that
`
`17 you can't use an index to find --
`
`18 Q Okay.
`
`19 A -- the attribute.
`
`20 Q I'm just asking you a question
`
`21 about how Kellogg works. I'm not talking
`
`22 about the claims, okay?
`
`23 So do you agree that the
`
`24 centroid of the object does not exist as
`
`25 a standalone variable that the database
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 33 of 187
`
`
`
`1 can access without using the object or
`
`2 the index?
`
`3 MR. MARTINELLI: Objection,
`
`4 form.
`
`5 A I don't agree.
`
`6 Q Why not?
`
`7 A Well, first, we're talking
`
`8 about just a particular implementation,
`
`9 and it specifically says that the index
`
`10 stores the centroids. So you have to --
`
`11 if they're stored somewhere you have to
`
`12 be able to go get them where they're
`
`13 stored. And whether you call that an
`
`14 index or something else I don't think is
`
`15 a matter of the claims of the patent.
`
`16 Q Again, I'm not asking about the
`
`17 claims of the patent. I'm asking about
`
`18 the functionality of Kellogg.
`
`19 In this example you pointed me
`
`20 to, at 4.3.2, if the system wants to
`
`21 access the attribute centroid of an
`
`22 object it must either do so through the
`
`23 index or through the object; correct?
`
`24 A It can go to, in this case, the
`
`25 index, and access information about the
`
`AVIGILON EX. 2018
`IPR2019-00311
`Page 34 of 187
`
`
`
`1 centroid, which would
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
