Case 1:14-cv-02396-PGG-SN Document 239-10 Filed 11/12/20 Page 1 of 6
`
`Exhibit 28
`
`

`

`Case 1:14-cv-02396-PGG-SN Document 239-10 Filed 11/12/20 Page 2 of 6
`
`Page 1
`
`Page 2
`
` A P P E A R A N C E S:
`
` RUSS AUGUST & KABAT
` Attorneys for Plaintiff
` 12424 Wilshire Boulevard, 12th floor
` Los Angeles, California 900
` BY: BRIAN D. LEDHAL, ESQ.
`
` SKADDEN ARPS SLATE MEAGHER & FLOM LLP
` Attorneys for Defendants
` 90 Park Avenue
` New York, New York 10036
` BY: DOUGLAS R. NEMEC, ESQ.
` ANDREW GISH, ESQ.
`
` ALSO PRESENT:
` JAMES ROBERTS, Legal video specialist
`
` - - -
`
`Page 4
`
` versus Google, Incorporated and YouTube
` LLC.
` The case is held in the U.S.
` District Court, Southern District of
` New York, Case No. 1:14-cv-02396-PGG.
` The name of the witness is
` Ingemar J. Cox.
` Counsel will, please, state their
` appearances for the record.
` MR. NEMEC: Douglas Nemec and
` Andrew Gish from Skadden Arps for the
` Defendants, Google and YouTube.
` MR. LEDAHL: And Brian Ledahl of
` Russ August & Kabat on behalf of the
` Plaintiff, Network-1.
` THE VIDEOGRAPHER: Our court
` reporter, Mayleen Ahmed, also of
` Veritext will please swear in the
` witness.
` _ _ _
` I N G E M A R J. C O X,
` called as a witness, having been duly
` sworn by a Notary Public, was examined
` and testified as follows:
`
`1 (Pages 1 to 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
`
`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
`
` UNITED STATES DISTRICT COURT
` SOUTHERN DISTRICT OF NEW YORK
` ---------------------------------X
` NETWORK-1 TECHNOLOGIES, INC., )
` )
` Plaintiff, )
` ) Civ. No.
` vs. ) 1:14-cv-02396
` ) (PGG)
` GOOGLE, INC. and YOUTUBE, LLC, )
` )
` Defendants. )
` ---------------------------------X
` June 12, 2015
` 9:00 a.m.
`
` * C O N F I D E N T I A L *
` UNDER THE PROTECTIVE ORDER
`
` VIDEOTAPED DEPOSITION OF
` INGEMAR J. COX, Ph.D., taken by Defendants,
` held at the offices of Amster Rothstein &
` Ebenstein LLP, 90 Park Avenue, New York, New
` York, pursuant to Notice, before Mayleen
` Cintrsn Ahmed a Registered Merit Reporter,
` Certified Realtime Reporter, and Notary
` Public of the State of New York.
`
`Job No. CS2079659
`
`Page 3
`
` THE VIDEOGRAPHER: Good morning.
` We are now on the record.
` Please note that the microphones
` are sensitive and may pick up
` whispering and private conversations.
` Please turn off all cell phones or
` place them away from the microphones as
` they can interfere with the deposition
` audio.
` Recording will continue until all
` parties agree to go off the record.
` My name is Jim Roberts
` representing Veritext Corporate
` Services with offices in New York City,
` New York.
`Today's date is June 12, 2015.
` The time is approximately 9:00 a.m.
` The deposition is being held at
` Amster Rothstein & Ebenstein located at
` 90 Park Avenue, New York City, New York
` and is being taken by counsel for the
` Defendant.
` The caption of the case is
` Network-1 Technologies, Incorporated
`
`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
`
`

`

`Case 1:14-cv-02396-PGG-SN Document 239-10 Filed 11/12/20 Page 3 of 6
`
`Page 65
`
`Page 66
`
` some vague recollections, but I -- I don't
` want to guess.
` Q. Do you have a recollection of what
` these enumerated types of feature extraction
` techniques were generally used for prior to
` your patent application?
` A. I think they could be used for a
` range of possibilities. So no, nothing
` specific.
` Q. You're not aware of any specifics
` within that range of possibility?
` A. Well, for example, you know,
` something like a discrete cosine transform,
` you know, would be used for MPEG compression,
` for example.
` Q. Okay. Have you ever used discrete
` cosine transformation for MPEG transmission
` in your work prior to your invention?
` A. I believe we did in the context of
` digital watermarking.
` Q. Had you ever used principal
` component analysis for feature extraction in
` your work at -- in your work, in your work
` anywhere prior to your invention?
`
`Page 68
`
` Q. Do you understand what's -- what's
` meant by "recognition literature" there?
` A. Well, I'll be thinking of
` literature in the scientific domain that
` describes, you know, ways to, to recognize
` objects or sounds or video.
` Q. So that -- that recognition
` literature, would that include pattern
` recognition?
` A. I think it would, yes.
` Q. And content recognition as well?
` A. Yes.
` Q. Is there a difference in your mind
` between content recognition and pattern
` recognition?
` A. Not a strong one. I mean, I'm
` not -- I haven't given it a lot of
` consideration. I mean, I have to give that
` some thought, I think, if there was a -- was
` or was not a distinction.
` Q. Would there at least be some
` overlap between pattern recognition and
` content recognition?
` A. Probably.
`
`17 (Pages 65 to 68)
`
`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
`
`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
`
` extraction, correct?
` A. Yes.
` Q. And the examples that are
` enumerated, for example, between about
` lines 20 and 43, those are techniques for
` feature extraction that were known prior to
` your invention, correct?
` A. About between 20 and 43, you said?
` Q. Roughly.
` A. I believe that this represents a
` sort of general description of -- of what
` features could, could be and what have been
` used in the past, yes.
` Q. All right. Had you personally
` used any of those enumerated feature
` extraction techniques prior to filing your
` patent application?
` A. I may have done in, in the context
` of, for example, a Pic Hunter work.
` Q. Do you remember what, what type of
` feature extraction technique you used in
` Pic Hunter?
` A. No. The short answer is I have to
` look at the Pic Hunter paper. I mean, I have
`
`Page 67
`
` A. I can't remember, but I have a
` vague feeling that I may have done.
` Q. Do you recall ever having used
` Fourier frequency decompositions in
` connection with feature extraction prior to
` your invention?
` A. I'm almost certain that I would
` have done, but I don't have any specific
` recollection of where I used it. It's a very
` common technique.
` Q. In connection with your invention,
` did you develop any new techniques for
` feature extraction?
` A. So, I don't -- I don't think I
` describe any, any new techniques for, for
` feature extraction.
` Q. In line -- it's around line 32 in
` column 7, there's a reference to recognition
` literature.
` (Witness perusing document.)
` A. Line which?
` Q. Around line 32.
` A. Okay. "The recognition literature
` contains..."
`
`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
`
`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
`
`

`

`Case 1:14-cv-02396-PGG-SN Document 239-10 Filed 11/12/20 Page 4 of 6
`
`Page 77
`
`Page 78
`
` A. Yes, I think it does.
` Q. What specific neighbor searching
` techniques are called out there?
` A. Well, I'm just reading from the
` paragraph here. But I mean, it says a number
` of possible data structures are applicable,
` kd-trees, vantage point trees, excluded
` middle vantage point forest. Those are the
` only ones I, I see at the moment.
` Q. And you understand those to be
` examples of neighbor searching for use in
` connection with your invention, correct?
` A. Well, I'm -- I'm hesitating due
` just to, to know whether there's specific
` examples of a nearest neighbor or a neighbor,
` and I can't remember, you know, which, which
` ones are which at the moment.
` Q. Okay. But do you, do you recall
` ever using kd-trees for either neighbor
` searching or nearest neighbor searching prior
` to your invention?
` A. Not that I can remember.
` Q. Do you recall ever using vantage
` point trees for either neighbor searching or
`
`Page 80
`
` considered to be a neighbor. Excuse me.
` One of those points will be
` closest, so that will be the nearest
` neighbor. But the other, other points that
` are slightly further away perhaps are still
` neighbors, it's just not the nearest
` neighbor.
` Q. So your -- just to make sure I
` understand.
` Your hesitation over whether
` kd-trees and vantage point trees are
` techniques for neighbor searching is that you
` don't recall whether they're used to identify
` that nearest point or to identify the others
` that may be within the circle but further out
` than that nearest point?
` A. That's correct.
` Q. And there's some descriptions
` starting at line 24 of column 22 in the '988
` of an excluded middle vantage point forest.
` Do you see that?
` A. Again? Column 24, line?
` Q. Column 22, line 24.
` A. I'm sorry. Line 24.
`
`20 (Pages 77 to 80)
`
`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
`
`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. I can't remember. I'm sorry.
` Q. Take a look, if you would, at
` column 22 in the '988 patent.
` Does column 2 describe any
` technique for neighbor searching within the
` context of your application, your patent?
` MR. LEDAHL: I'm sorry. 2 or 22?
` MR. NEMEC: I meant to say 22. I
` may have said 2.
` MR. LEDAHL: That's okay. I just
` wanted to make sure.
` A. And the question again, please?
` Q. Does column 22 describe techniques
` for neighbor searching within the context of
` your invention?
` A. Let me just, again, take a look at
` it.
` (Witness perusing document.)
` A. And one more time, the question?
` Sorry.
` Q. Does the text in that column
` describe examples of neighbor searching
` techniques within the context of your
` invention?
`
`Page 79
`
` nearest neighbor searching prior to your
` invention?
` A. Not that I can remember.
` Q. Since, since you made a
` distinction here, I should re-ask a question
` that I posed earlier differently.
` Prior to your invention, had you
` ever used nearest neighbor searching in any
` of your research?
` A. Again, I'm not sure. But -- I'm
` not sure.
` Q. And what, what --
` Why is it that you're drawing a
` distinction between neighbor searching and
` nearest neighbor searching here when I'm
` asking about kd-trees and vantage point
` trees?
` A. Well, again, going back to the
` example of a reference point and then other
` points. So, you know, if you're given, for
` example, this threshold that we talked about,
` you can image that you have a -- have a
` reference point and a circle around it. And
` so any points within that circle are
`
`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
`
`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
`
`

`

`Case 1:14-cv-02396-PGG-SN Document 239-10 Filed 11/12/20 Page 5 of 6
`
`Page 81
`
`Page 82
`
` mind if we take a break?
` Q. Sure. That will be fine. This is
` a good time.
` THE VIDEOGRAPHER: Going off the
` record at 10:36 a.m. This is the end
` of disc one in the deposition of
` Ingemar J. Cox.
` (Whereupon, a short recess was
` taken.)
` THE VIDEOGRAPHER: Going back on
` the record. 10:48 a.m. This is the
` beginning of disc two in the deposition
` of Ingemar J. Cox.
` BY MR. NEMEC:
` Q. Dr. Cox, before the break we were
` talking about some of the terminology that
` appears in your patents, and I want to turn
` to another term.
` The term "non-exhaustive search"
` is used in the context of your patents,
` correct?
` A. Yes.
` Q. For example, in the '988 patent,
` the term "non-exhaustive search" appears in
`
`Page 84
`
` documents.
` And even that same algorithm, I
` mean, it might be that you could stop before
` you see, see the end in that if you found an
` identical document, you know, at that point,
` you know, it would not be -- no point in
` trying to look for something more similar
` than an identical one. So you could stop.
` But still, in principle the algorithm would
` be, you know, exhaustive in that, you know,
` you would have to look at all of documents.
` In contrast, you know, a
` non-exhaustive example, example might be one
` where the documents are in piles where -- I
` don't know. Perhaps they're given in their
` title. So the A documents in one pile, the
` B documents, et cetera.
` Now you're given a query with a
` document that starts with the letter C and,
` you know, because of that, you're able to
` just look at the documents in pile C so, so
` you would never look at all of the documents
` in the database. So that would be a
` non-exhaustive search.
`
`21 (Pages 81 to 84)
`
`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
`
`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
`
` (Witness perusing document.)
` A. Yes. I see a reference to
` excluded middle vantage point forest, yes.
` Q. And do you understand excluded
` middle vantage point forest to be a technique
` for neighbor searching in the context of your
` invention?
` A. I can't remember. I can't
` remember.
` Q. Do you know if it's a technique
` for nearest neighbor searching?
` A. I haven't looked at these, these
` algorithms for years, so I'm just -- don't
` want to make a mistake and say one is one and
` another is another.
` Q. Okay. Do you propose in your --
` in your patent, in the '988 patent, any
` techniques for neighbor searching that were
` not known prior to your invention?
` A. I don't believe that I did, no.
` Q. Do you pose any techniques for
` nearest neighbor searching that were not
` known prior to your invention?
` A. I don't remember doing so. Do you
`
`Page 83
`
` Claim 15, correct?
` A. So we're on Exhibit 17?
` Q. We are, yes.
` A. And we're on --
` Q. I gave Claim 15 as an example.
` (Witness perusing document.)
` A. Yes. I see that term in there.
` Q. What's your understanding of the
` term "non-exhaustive search" as it's used in
` your patents?
` A. So maybe it should be easier to
` start with what I mean by -- what's meant by
` an exhaustive search.
` So, for example, if you have a,
` say, a series of documents and they're just
` all, all randomly on the table and you're
` asked -- you're given a document and you're
` asked to find the most similar document,
` then, you know, the al -- the algorithm that
` you would use would presumably be that you
` can pair it with the first document on the
` table, then the second, then the third, and
` look at all -- all the documents. So that
` would be an exhaustive look at all 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
`
`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
`
`

`

`Case 1:14-cv-02396-PGG-SN Document 239-10 Filed 11/12/20 Page 6 of 6
`
`Page 85
`
`Page 86
`
` publications.
` Q. You don't recall specifically
` having used the term "non-exhaustive" in your
` writings prior to your patents?
` A. I don't recall, no.
` Q. Was the concept of non-exhaustive
` searching known prior to your patent
` application?
` A. I believe it was, yes.
` Q. Do you know in what context
` non-exhaustive searching was, was used prior
` to your invention?
` A. What do you mean by a context
` here?
` Q. Are you aware of applications for
` which non-exhaustive searching was used prior
` to your invention?
` A. I would say more, you know, just
` the general field of search. You know, I was
` familiar with sort of the concept of
` non-exhaustive search.
` Q. Had you used non-exhaustive
` searching techniques in any of your work
` prior to your inventions?
`
`Page 88
`
` can you -- can you repeat the question,
` though?
` Q. Sure. The question was whether
` your patent discloses examples of
` non-exhaustive search techniques?
` A. Okay. So I mean, I think --
` sorry. I was sort of -- didn't fully hear
` the question the first time.
` Q. That's okay.
` A. So I think in many of the -- of
` the examples that we've referred to already,
` you know, in some circumstances, you know,
` and maybe in many circumstances, many of the
` search techniques that are listed here would
` be non-exhaustive.
` Q. Where in particular are you
` pointing?
` A. Well, as an example, you know,
` column 22, the top of column 22 is a number
` of, of data structures, kd-tree, vantage
` point trees, excluded middle trees which
` under some circumstances may, may be
` non-exhaustive, in other -- in other
` circumstances, they may revert to being
`
`22 (Pages 85 to 88)
`
`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
`
`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
`
` Q. Are you aware of other ways to
` describe a non-exhaustive search than the
` example you just used?
` MR. LEDAHL: Vague and ambiguous.
` A. I'm sure there are other ways, but
` I -- that's the best I can provide you with
` at the moment.
` Q. At the time that you filed your
` original patent application, do you believe
` that the meaning of "non-exhaustive" as
` you've just described it was well-known to
` people in your field?
` A. I think it was, yes.
` Q. Do you think others might have
` described it as a somewhat different way than
` you did at that time?
` MR. LEDAHL: Calls for
` speculation.
` A. I wouldn't be able to answer that.
` Q. Have you ever used the term
` "non-exhaustive" in any of your writings
` prior to your patents?
` A. Again, I wouldn't know. I would
` have to, literally have to do a search of my
`
`Page 87
`
` A. Again, I can't remember. One
` possibility might have been Pic Hunter.
` Q. Are you aware of any of your peers
` at NEC having used non-exhaustive searching
` in any of the projects they worked on?
` A. Well, I would imagine that sort of
` PD Yianilos, you know, was -- who was doing
` work on the search would have, would have
` done that.
` Q. Do you know what, for what purpose
` Mr. Yianilos would have been doing that?
` THE VIDEOGRAPHER: Counsel, I'm
` sorry. You're covering your
` microphone.
` MR. NEMEC: Oh, I apologize. Did
` you catch the question?
` THE VIDEOGRAPHER: Yes.
` A. I'm sorry. I don't remember what
` PD Yianilos was specifically working on at
` that time.
` Q. Does your patent disclose examples
` of non-exhaustive search techniques?
` A. I don't know. I mean, I haven't
` read the patents in, in a long time. Sorry,
`
`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
`
`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
`
`