`Case 1:14-cv-02396—PGG-MHD Document 158-2 Filed 07/19/19 Page 1 of 35
`
`
`
`
`
`
`
`
`
`
`
`EXHIBIT 1
`
`
`
`EXHIBIT 1
`
`
`
`Case 1:14-cv-02396-PGG-MHD Document 158-2 Filed 07/19/19 Page 2 of 35
`JAMES A. STORER, PH.D.
`NETWORK-1 TECHNOLOGIES, INC. v.
`July 10, 2019
`GOOGLE, LLC and YOUTUBE, LLC
`Page 3
`
`Page 1
`
` 1 UNITED STATES DISTRICT COURT
`
` 2 SOUTHERN DISTRICT OF NEW YORK
`
` 3 - - - - - - - - - - - - - - - -x
`
` 4 NETWORK-1 TECHNOLOGIES, :
`
` 5 INC. :
`
` 6 Plaintiff : 14 Civ. 2396(PGG)
`
` 7 v. :
`
` 8 GOOGLE, LLC and YOUTUBE, : 14 Civ. 9558(PGG)
`
` 9 LLC :
`
`10 Defendants :
`
`11 - - - - - - - - - - - - - - - -x
`
`12
`
`13
`
`14
`
`15 VIDEOTAPED DEPOSITION OF
`
`16 JAMES A. STORER, PH.D.
`
`17 Washington, D.C.
`
`18 Wednesday, July 10, 2019
`
`19 8:11 a.m.
`
`20
`
`21
`
`22
`
`23 Reported By: Joan V. Cain
`
`24
`
`25
`
` 1 A P P E A R A N C E S
`
` 2
`
` 3 ON BEHALF OF PLAINTIFF:
`
` 4 MARC A. FENSTER, ESQUIRE
`
` 5 RUSS AUGUST & KABAT
`
` 6 12424 Wilshire Boulevard
`
` 7 Twelfth Floor
`
` 8 Los Angeles, California 90025
`
` 9 Telephone: (310) 826-7474
`
`10 E-mail: mfenster@raklaw.com
`
`11
`
`12 ON BEHALF OF DEFENDANTS:
`
`13 GRAHAM SAFTY, ESQUIRE
`
`14 ANDREW V. TRASK, ESQUIRE
`
`15 SAMUEL BRYANT DAVIDOFF, ESQUIRE
`
`16 WILLIAMS & CONNOLLY, LLP
`
`17 725 Twelfth Street, Northwest
`
`18 Washington, D.C. 20005
`
`19 Telephone: (202) 434-5000
`
`20 E-mail: gsafty@wc.com
`
`21 atrask@wc.com
`
`22 sdavidoff@wc.com
`
`23
`
`24
`
`25
`
`Page 2
` 1 Videotaped Deposition of JAMES A. STORER,
`
` 2 Ph.D., held at the law offices of:
`
` 3
`
` 4 WILLIAMS & CONNOLLY, LLP
`
` 5 725 Twelfth Street, Northwest
`
` 6 Washington, D.C. 20005
`
` 7 (202) 434-5000
`
` 8
`
` 9 Pursuant to Notice, before Joan V. Cain,
`
`10 Court Reporter and Notary Public in and for the
`
`11 District of Columbia.
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
` 1 A P P E A R A N C E S C O N T I N U E D
` 2
` 3 ALSO PRESENT:
` 4 Malcolm Peplow, Videographer
` 5 Andrew Borrasso, Summer Associate
` 6 Williams & Connolly
` 7
` 8
` 9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`Page 4
`
`Min-U-Script®
`
`Barkley Court Reporters
`
`(1) Pages 1 - 4
`
`
`
`Case 1:14-cv-02396-PGG-MHD Document 158-2 Filed 07/19/19 Page 3 of 35
`JAMES A. STORER, PH.D.
`NETWORK-1 TECHNOLOGIES, INC. v.
`July 10, 2019
`GOOGLE, LLC and YOUTUBE, LLC
`Page 7
`
`Page 5
`
` 1 C O N T E N T S
`
` 2 EXAMINATION OF JAMES A. STORER, PH.D. PAGE
`
` 3 By Mr. Fenster 7
`
` 4
`
` 5 E X H I B I T S
`
` 6 (Attached to the Transcript.)
`
` 7 STORER DEPOSITION EXHIBITS PAGE
`
` 8 EXHIBIT 1 Declaration of Professor 28
`
` 9 James A Storer, 6/28/19
`
`10 EXHIBIT 2 Dr. Storer's Curriculum 16
`
`11 Vitae
`
`12 EXHIBIT 3 U.S. Court of Appeals for 31
`
`13 the Federal Circuit Decision
`
`14 in Google LLC, v. Network-1
`
`15 Technologies, Inc., 3/26/18
`
`16 EXHIBIT 4 U.S. Court of Appeals for 56
`
`17 the Federal Circuit Decision
`
`18 in Google, LLC v. Network-1
`
`19 Technologies, Inc., 1/23/18
`
`20
`
`21
`
`22
`
`23
`
`24
`
`25
`
` 1 proceed.
` 2 EXAMINATION BY COUNSEL FOR PLAINTIFF
` 3 BY MR. FENSTER:
` 4 Q Good morning, Dr. Storer.
` 5 A Good morning.
` 6 Q How are you today?
` 7 A Good.
` 8 Q Anything that would prevent you from giving
` 9 your best testimony today?
`10 A No.
`11 Q How'd you come to be retained in this case?
`12 A Someone called me.
`13 Q Who?
`14 A You know, I don't remember.
`15 Q Was it someone from the law firm of
`16 Williams & Connolly?
`17 A Yes.
`18 Q Okay. Do you recall approximately when you
`19 were called?
`20 A I think on the order of a year ago is when
`21 I first was approached.
`22 Q And what were you asked to do?
`23 A The work I've done has involved claim
`24 construction.
`25 Q Have you done any work other than related
`
`Page 6
`
`Page 8
`
` 1 P R O C E E D I N G S
` 2 THE VIDEOGRAPHER: Good morning. We are
` 3 going on the record at 8:11 a.m., on Wednesday, July
` 4 10th, 2019. This is Media Unit 1 of the video
` 5 recorded deposition of James Storer.
` 6 Counsel and all present in the room will
` 7 now state their appearances and affiliations for the
` 8 record.
` 9 MR. FENSTER: Marc Fenster with Russ August
`10 Kabat on behalf of the Plaintiff Network-1
`11 Technologies, Inc.
`12 MR. SAFTY: Graham Safty of Williams &
`13 Connolly on behalf of Defendants Google, LLC, and
`14 YouTube, LLC. Also with me today is Sam Davidoff,
`15 partner at Williams & Connolly on behalf of
`16 defendants, as well as Andrew Borrasso, a summer
`17 associate at Williams & Connolly on behalf of
`18 defendants.
`19 THE VIDEOGRAPHER: Will the court reporter
`20 please swear in the witness.
`21 JAMES A. STORER, PH.D.
`22 having been duly sworn under penalties of perjury by
`23 the Notary Public, was examined and did testify as
`24 follows:
`25 THE VIDEOGRAPHER: Thank you. We may
`
` 1 to claim construction?
` 2 A No.
` 3 Q Have you talked to anyone at Google about
` 4 the Content ID system?
` 5 A No.
` 6 Q Have you talked to anyone at YouTube about
` 7 the Content ID system?
` 8 A No.
` 9 Q Have you talked to anyone at Google or
`10 YouTube about anything related to this case?
`11 A No.
`12 Q Have you previously worked for clients of
`13 Williams & Connolly?
`14 A I don't believe so. Well, I don't know who
`15 all their clients are, so I don't know whether
`16 someone I had -- I'm not sure what you mean by
`17 worked for either. But if you meant that have I
`18 been retained in a case where the clients were
`19 also -- have also been clients of Williams &
`20 Connolly, but maybe they were clients of someone
`21 else, I can't say for sure in that sense, but I
`22 don't have any rec- -- specific recollection of
`23 working with this firm before.
`24 Q Okay. Have you ever done any -- have you
`25 ever -- strike that.
`
`Min-U-Script®
`
`Barkley Court Reporters
`
`(2) Pages 5 - 8
`
`
`
`Case 1:14-cv-02396-PGG-MHD Document 158-2 Filed 07/19/19 Page 4 of 35
`JAMES A. STORER, PH.D.
`NETWORK-1 TECHNOLOGIES, INC. v.
`July 10, 2019
`GOOGLE, LLC and YOUTUBE, LLC
`Page 11
`
`Page 9
`
` 1 Have you ever been retained to work on a
` 2 case involving Google or YouTube?
` 3 A In the past there have been a few large
` 4 joint defenses, and I don't know whether Google
` 5 might have been one of the defendants in the large
` 6 group, but they wouldn't have been specifically the
` 7 one I was -- I don't have any specific recollection
` 8 of specifically working on behalf of Google.
` 9 Q So to your knowledge, you've never been
`10 retained by or on behalf of Google or YouTube in
`11 connection with a litigation?
`12 A That's my best recollection sitting here,
`13 yes.
`14 Q Have you ever -- outside the litigation
`15 context, have you ever done any work or consulting
`16 for Google or YouTube?
`17 A I had a grant a while back from Google,
`18 quite a while back. It went to support some work by
`19 a graduate student.
`20 Q Anything else?
`21 A No.
`22 Q You never had any funding -- you've never
`23 had any research funded by Google except for this
`24 grant that you just referred to?
`25 A There's a small grant that they give for,
`
` 1 yeah, students. Yeah, nothing beyond that, a number
` 2 of -- maybe -- I don't know -- eight years ago or
` 3 something.
` 4 Q And who was the student?
` 5 A The student was Kevin Thomas.
` 6 Q And do you know the amount of the grant?
` 7 A It's relatively small. There's a --
` 8 there's a program they had. I wish I remembered the
` 9 name of the program. I don't remember the details.
`10 It's been a long time. Sorry.
`11 Q Any idea what the order of magnitude of the
`12 grant was?
`13 A Less than $100,000 my best recollection.
`14 Q And to your -- to your recollection, that
`15 grant eight years ago in connection with Kevin
`16 Thomas is the only Google or YouTube funding that
`17 you've ever received in connection with any work
`18 that you've been involved in?
`19 A Correct.
`20 Q And you are a professor at Brandeis; is
`21 that correct?
`22 A Yes.
`23 Q And what department are you in there?
`24 A Computer science department.
`25 Q And are you aware whether the computer
`
` 1 science department at Brandeis has received any
` 2 funding from Google or YouTube?
` 3 A I don't recall anything -- I don't recall
` 4 anything other than the small grant that I had
` 5 myself. Not that I know of, no.
` 6 Q Has your department applied for grants from
` 7 Google or YouTube, other than the one that was
` 8 granted, that have not been granted?
` 9 A So I'll try to answer your question more
`10 broadly. Generally speaking, it wouldn't be the --
`11 exactly the -- usually it's a faculty member who
`12 applies, not formally -- I guess the university
`13 would sign off on the application, but it would be
`14 sort of a research group applying for a grant is how
`15 it works, not so much the department. You could
`16 have the department applying for some big -- big,
`17 department-wide grant.
`18 Surely nothing like that that I know of,
`19 Google, and I don't know of any particular faculty
`20 member who has, but I can't preclude it either
`21 because I'm not always involved in that.
`22 Q Have you applied to any -- applied to
`23 Google or YouTube for any funding other than in
`24 connection with that one for Kevin Thomas?
`25 A No.
`
`Page 10
`
`Page 12
`
` 1 Q Have you been asked to analyze any issues
` 2 related to infringement in this case?
` 3 A No. My work has been limited to claim
` 4 construction.
` 5 Q Other than indefiniteness that you discuss
` 6 in your declaration, have you done any work on
` 7 validity?
` 8 A I'm not sure exactly how you're
` 9 characterizing my declaration, but putting that
`10 aside, I've not done any work on validity.
`11 Q And have you reviewed any depositions in
`12 this case?
`13 A Yes.
`14 Q What depos have you read?
`15 A I read the deposition of Dr. Mitzenmacher.
`16 Q Any others?
`17 A Not that I recall.
`18 Q Did you review any inventor depositions in
`19 this case?
`20 A No, not that I recall.
`21 Q Do you know Dr. Mitzenmacher?
`22 A Yes.
`23 Q How do you know Dr. Mitzenmacher?
`24 A You know, I don't know him well, but quite
`25 a few years ago he came out for lunch at the time
`
`Min-U-Script®
`
`Barkley Court Reporters
`
`(3) Pages 9 - 12
`
`
`
`Case 1:14-cv-02396-PGG-MHD Document 158-2 Filed 07/19/19 Page 5 of 35
`JAMES A. STORER, PH.D.
`NETWORK-1 TECHNOLOGIES, INC. v.
`July 10, 2019
`GOOGLE, LLC and YOUTUBE, LLC
`Page 15
`
`Page 13
`
` 1 that I was finishing my book, and we had a nice day
` 2 talking about research, and I don't know if I've met
` 3 him in person since then, but know of him.
` 4 Q Do you have any opinions of him as a
` 5 researcher or scientist, lecturer?
` 6 A He's a nice guy. I have no -- nothing -- I
` 7 don't have anything bad to say about him if that's
` 8 what you're asking. Nice guy. I enjoyed meeting
` 9 him.
`10 Q Have you read any of his papers?
`11 A Not recently, but in the past, yes.
`12 Q Have you read anything else that he's
`13 written?
`14 A Well, we mentioned the -- his
`15 declaration -- or he had a -- his deposition, his
`16 declaration. Oh, and one other thing too. He was
`17 on the opposite side in a different case I was on.
`18 So I would have read documents he wrote in that
`19 case.
`20 Q And what case was that?
`21 A That was a case of -- of Realtime versus
`22 Oracle I believe. That's my best memory.
`23 Q Have you ever attended any lectures by
`24 Dr. Mitzenmacher or seen any talks by him?
`25 A So the day that he visited me, I believe he
`
` 1 gave a talk. It was a long time ago, but I believe
` 2 so -- I believe so.
` 3 Q Do you know where that talk would have
` 4 been?
` 5 A In my department, yeah. Again, that's my
` 6 best recollection. It may be he just visited, but I
` 7 kind of think he gave a talk. I think I would
` 8 normally -- someone's going to visit, have him give
` 9 a talk. That's my best recollection, but it's been
`10 a long time.
`11 Q Did you organize his visit to Brandeis?
`12 A My best recollection is -- I'm not sure if
`13 I organized it. It might be that I suggested it to
`14 our seminar series person. I don't actually
`15 remember the details. But my best recollection or
`16 best estimate of what I would have done is that
`17 maybe I would have wanted to have him come out and
`18 suggested he be invited as one of the lecturers on
`19 the weekly lecture thing and then have lunch with
`20 him, that sort of thing.
`21 Again, it's -- I'm not positive, but you
`22 could ask him as well, but that's my best
`23 recollection.
`24 Q What is this program where you invite
`25 weekly lecturers or invite lecturers on a weekly
`
` 1 basis?
` 2 A Not a program, the department just
` 3 periodically has speakers come, and usually someone
` 4 might be in charge of doing that. Different years
` 5 it's worked different ways. Some years it might be
` 6 the chair in charge. Some years some faculty might
` 7 be organizing periodic seminars. Sometimes people
` 8 just invite people to talk to their research group
` 9 but then invite others to -- to come. It's not
`10 super formal actually.
`11 Q Do you know who Dr. Mitzenmacher spoke to?
`12 Was it a department? A group? A class?
`13 A No. Oh, who -- I don't know -- again, I
`14 don't remember even definitely that he did give a
`15 talk, but if he did, it would have been anyone who
`16 wanted to come.
`17 Q Do you know what Dr. Mitzenmacher's area of
`18 research is or area of expertise is?
`19 A At the time I met him, we -- we discussed
`20 algorithms, and algorithms aside, I don't really
`21 know what he's been doing recently actually.
`22 Q Have you spoken with any other experts in
`23 this case?
`24 A No, not that I recall. No. I mean,
`25 related to this case obviously, yes. Yeah.
`
`Page 14
`
`Page 16
`
` 1 Q Do you agree that the subject matter of the
` 2 patents in this case relate to neighbor searches in
` 3 high-dimensional space?
` 4 A Well, searching for nearest neighbors is
` 5 something that's discussed in the specifications.
` 6 Q Have you written any papers involving
` 7 algorithms for searching for near or nearest
` 8 neighbors in high-dimensional space?
` 9 A Well, I'm not sure what you're encompassing
`10 when you say that, but I have had papers in the past
`11 and a couple Ph.D. theses that would have related to
`12 content-based image retrieval, which may fall under
`13 the umbrella of what you're asking.
`14 (Storer Deposition Exhibit 2 was
`15 marked for identification.)
`16 BY MR. FENSTER:
`17 Q I'll hand you what's been marked as Storer
`18 Exhibit 2. This is Exhibit A and your CV that was
`19 attached to your declaration; is that correct?
`20 A Well, it looks like a copy of my academic
`21 vitae, so I assume it's the copy that was attached
`22 to -- if you say so, that sounds perfectly
`23 reasonable.
`24 Q Is this CV accurate and up to date?
`25 A Let me take a look. It looks pretty --
`
`Min-U-Script®
`
`Barkley Court Reporters
`
`(4) Pages 13 - 16
`
`
`
`Case 1:14-cv-02396-PGG-MHD Document 158-2 Filed 07/19/19 Page 6 of 35
`JAMES A. STORER, PH.D.
`NETWORK-1 TECHNOLOGIES, INC. v.
`July 10, 2019
`GOOGLE, LLC and YOUTUBE, LLC
`Page 19
`
`Page 17
`
` 1 there may be one or two additional papers, deep
` 2 learning related papers. That's something my
` 3 students have been working on recently, but pretty
` 4 much it looks -- the publications look up to date,
` 5 more or less. Yeah, pretty -- pretty up-to-date
` 6 version.
` 7 Q Okay. You mentioned a few papers relating
` 8 to content-based image retrieval as may be relating
` 9 to neighbor searches in high-dimensional space. You
`10 recall that?
`11 A I recall a question about that, yes.
`12 Q Okay. Have you written any other papers
`13 other than those related to content-based image
`14 retrieval that relate to neighbor or near-neighbor
`15 searching in high-dimensional space?
`16 A Again, I'm not sure what you're
`17 encompassing in -- in that broad category. It could
`18 be that some people would include some -- include
`19 some of the deep learning work that's referenced as
`20 well. Again, it's hard to answer your question
`21 specifically because it's sort of a broad
`22 categorization.
`23 I guess maybe another answer you might
`24 consider would be some of the vector quantization
`25 work, where you are sometimes trying to perform
`
` 1 vector quantization and the encoder has to match
` 2 vectors that are -- that come from high-dimensional
` 3 space.
` 4 Q Have you ever designed an algorithm to do a
` 5 neighbor search in high-dimensional space?
` 6 A I think the answer would be pretty similar
` 7 to my previous answers in that the publications tend
` 8 to encompass the work that I've done, and you asked
` 9 about the publications and said the broad category,
`10 so I guess the answer would be pretty much similar
`11 to the previous answer.
`12 Q The question here was different. Have you
`13 designed -- have you ever designed an algorithm to
`14 do a neighbor search in high-dimensional space?
`15 A So, again, your question is broad, and when
`16 you asked about have I published in those areas, the
`17 publications, if you were to consider them to fall
`18 within the breadth of your question or not, would
`19 reflect the design of algorithms, for example, as
`20 you asked in your question. It was done possibly in
`21 conjunction with a co-author or a student.
`22 Q That's not really my question, Dr. Storer,
`23 and I'm not asking about what's in your publication.
`24 I'm asking: Have you personally ever designed an
`25 algorithm to do a neighbor search in
`
` 1 high-dimensional space?
` 2 A So I think I answered the question as best
` 3 I could. It's broad. It's not clear what you're
` 4 including or not, but to the extent -- for example,
` 5 you put the vitae in front of me. Those papers do
` 6 reflect research where algorithms have been
` 7 designed, possibly in conjunction with a student, so
` 8 that if you consider the topic of the paper to fall
` 9 within the category of your question, then, yes, I
`10 would have been doing that sort of design.
`11 Q Okay. I'm not asking about a paper that
`12 might discuss it. I'm asking about your design
`13 work. Have you personally designed an algorithm to
`14 do a neighbor search in high-dimensional space?
`15 MR. SAFTY: Objection, asked and answered.
`16 THE WITNESS: It seems to be the same
`17 question you've asked, and, again, I was saying your
`18 question's very broad. It's not clear what you are
`19 considering in or out, but as I did say in the
`20 answer, these papers, which I mentioned a number of
`21 categories, did involve actual design by myself and
`22 in conjunction with graduate students, but, again,
`23 you'd have to decide whether you felt the paper was
`24 within the scope of the broad category you've
`25 described or not.
`
`Page 18
`
`Page 20
`
` 1 BY MR. FENSTER:
` 2 Q What's unclear about a neighbor search in
` 3 high-dimensional space to you?
` 4 A I didn't say it was unclear. I just said
` 5 it was very broad, and I'm not sure what you're
` 6 considering to be within that category or not, and I
` 7 tried to broadly answer your question about a number
` 8 of different categories of papers that are in the
` 9 vitae you gave to me and --
`10 Q Do you know what a neighbor search is?
`11 A I know what it -- certainly have done --
`12 know what a neighbor search is in a number of
`13 different contexts.
`14 Q What's your understanding of neighbor
`15 search?
`16 MR. SAFTY: Objection, vague and ambiguous.
`17 THE WITNESS: Again, I think you're -- I
`18 guess I -- it might be easier to answer the
`19 question, for example, in the context of the patent
`20 at issue or my expert report. We're looking at
`21 particular kinds of issues in this case, and you
`22 seem to be asking a question that -- I'm not really
`23 sure the breadth of the question I guess is what I'm
`24 saying. The terminology is quite broad and is used
`25 in a lot of different contexts.
`
`Min-U-Script®
`
`Barkley Court Reporters
`
`(5) Pages 17 - 20
`
`
`
`Case 1:14-cv-02396-PGG-MHD Document 158-2 Filed 07/19/19 Page 7 of 35
`JAMES A. STORER, PH.D.
`NETWORK-1 TECHNOLOGIES, INC. v.
`July 10, 2019
`GOOGLE, LLC and YOUTUBE, LLC
`Page 23
`
`Page 21
`
` 1 BY MR. FENSTER:
` 2 Q Are you unclear about what high-dimensional
` 3 space means?
` 4 MR. SAFTY: Objection, argumentative.
` 5 THE WITNESS: I don't think I'm unclear
` 6 about anything, but I do -- I do want to be sure I
` 7 understand what you're encompassing in your
` 8 question. Even the term "high-dimensional space"
` 9 can refer to a lot of different kinds of spaces, so
`10 it's obviously a term that -- a phrase that you see
`11 in many contexts.
`12 BY MR. FENSTER:
`13 Q What's your understanding of
`14 high-dimensional space --
`15 A Could be --
`16 Q -- in the search context?
`17 A Could be in many different contexts. For
`18 example, one context -- it was the last one we
`19 touched on -- one, I mentioned yet another category
`20 that you might look to in my -- in my CV of vector
`21 quantization, and if you are now having vectors that
`22 have, let's say, a typical one might have anywhere
`23 from 16 to 64 components in it, maybe more, then
`24 that might be considered by some to be a
`25 high-dimensional space.
`
` 1 Q Can you identify an algorithm that you have
` 2 designed for searching for near neighbors in
` 3 high-dimensional space?
` 4 A And, again, your question seems very broad
` 5 in that exactly what you are encompassing in -- in
` 6 that category, but I could imagine, as we mentioned
` 7 earlier, that if you were to look at my vitae,
` 8 papers that related to the vector quantization we
` 9 just talked about, context-based image retrieval,
`10 and some of the deep learning papers would relate.
`11 And so, for example, going back to the one
`12 we just talked about, vector quantization, a number
`13 of -- a number of different -- a number of papers
`14 over a span of time in this vitae would have
`15 addressed work that I did in conjunction with a
`16 graduate student to do efficient searching for
`17 matching vectors in those spaces, again, if you
`18 consider a space that was -- for example,
`19 64-component vectors to be high dimensional.
`20 Q Okay. Can you identify one for me, please?
`21 A Sure. We can start way back at the
`22 beginning. So in the context of vector
`23 quantization -- go back -- well, I mean, again, I'm
`24 starting from the back working forward, but, for
`25 example, I remember one of the earliest students I
`
` 1 worked with -- with on the subject vector
` 2 quantization was a student Eric Lin; On the
` 3 Complexity of Optimal Tree Pruning for Source Coding
` 4 would have been something that dealt with vector
` 5 quantization.
` 6 Q Can you point me to where you are, please?
` 7 A Oh, I've started furthest in the past
` 8 working forward. It's titled "Page 16 of 19," the
` 9 first reference. I just happened to see it. Again,
`10 I may have skipped over some, but just to -- he --
`11 similar -- again, remembering his work, he had a --
`12 he did a Ph.D. thesis. There's another one on the
`13 following page: Resolution-Constrained
`14 Tree-Structured Vector Quantization for image
`15 compression.
`16 So, actually, again, depending on your --
`17 your scope, you might consider some of the video
`18 compression work to fall in that category that was
`19 done by another graduate student. I'm seeing a
`20 split/merge parallel block-matching algorithm.
`21 Again, as I move backward -- forward in time,
`22 another algorithm by Eric Lin on that same page
`23 moving forward in time, and so on.
`24 Q So you pointed out the paper by Eric Lin at
`25 the top of page 16 of 19 entitled "On the Complexity
`
`Page 22
`
`Page 24
`
` 1 of Optimal Tree Pruning for Source Coding."
` 2 Do you see that?
` 3 A Yes. And then a couple other papers by him
` 4 on the following page, for example, yes.
` 5 Q Did you design an algorithm for
` 6 near-neighbor searching in high-dimensional space
` 7 that's described in that paper?
` 8 A So, again, I'm not sure what you're
` 9 encompassing with the broad term, but I can tell you
`10 what he did in that paper, and you can decide for
`11 yourself.
`12 Q I'm asking -- I'm trying to find what you
`13 designed.
`14 A So I can say what he designed in that
`15 paper, if you like.
`16 Q Okay. I'm not asking what he designed.
`17 I'm asking what you have --
`18 A The two of us, yeah.
`19 Q -- ever designed.
`20 A We worked together in this. We were a
`21 Ph.D. student and advisor working together and
`22 co-authoring the papers, and, for example, in that
`23 work -- we had -- he had a number of papers. But
`24 the theme of the thesis was that in this case you
`25 have vectors which could be as -- could be as small
`
`Min-U-Script®
`
`Barkley Court Reporters
`
`(6) Pages 21 - 24
`
`
`
`Case 1:14-cv-02396-PGG-MHD Document 158-2 Filed 07/19/19 Page 8 of 35
`JAMES A. STORER, PH.D.
`NETWORK-1 TECHNOLOGIES, INC. v.
`July 10, 2019
`GOOGLE, LLC and YOUTUBE, LLC
`Page 27
`
`Page 25
`
` 1 as possibly four, but typically might be more in the
` 2 range of 16 to 64, could be bigger, and you're
` 3 trying to do matching, nearest-neighbor matching, of
` 4 new material to a -- in this case, a trained and --
` 5 and statically stored, if you will, dictionary of --
` 6 or database of these vectors.
` 7 Later work we hadn't got to yet would look
` 8 at adaptive methods for that, but in this case a
` 9 pretrained database, but, on the other hand, he's
`10 also concerned with tuning that database to enable
`11 the subsequent searching of the database to be
`12 faster. I guess that's one way to characterize his
`13 work.
`14 Q Can you identify any algorithms that you've
`15 written by yourself without help for [sic] others
`16 for searching for neighbors in high-dimensional
`17 space?
`18 A I'm not sure how to divide the work when I
`19 work. Generally, in academia you generally work
`20 together with graduate students. Often, in many
`21 cases there are papers where maybe I designed the
`22 algorithm; graduate student implemented. It could
`23 be we worked together to design the algorithm; could
`24 be the graduate student designed the initial, and I
`25 suggested improvements. It could be all possible
`
` 1 combinations.
` 2 Q And because of that, you can't identify any
` 3 algorithms that you've written by yourself without
` 4 help for others for searching -- for searching for
` 5 neighbors in high-dimensional space?
` 6 MR. SAFTY: Objection, misstates the
` 7 testimony.
` 8 THE WITNESS: Actually, I would agree with
` 9 the objection that was just made. I don't think
`10 that's what I said at all. I think in many cases --
`11 in fact, I'm thinking of another graduate student --
`12 actually, quite a few of these graduate students.
`13 So I'm thinking of another graduate student
`14 where really initially, I had -- I had to solely
`15 design the algorithm, but I had him perform
`16 experiments, and after he performed experiments, we
`17 were able to improve it, and he actually ended up
`18 having a nice experimental thesis, for example, and
`19 went on to have a very successful career at IBM.
`20 BY MR. FENSTER:
`21 Q And who is that?
`22 A That's Cornel Continescu. It's a hard name
`23 to pronounce. That's later on in time than the
`24 ones -- we're starting from the oldest to newest, so
`25 that's later on in time. He was looking at -- it's
`
` 1 something that I had been interested in for a long
` 2 time and that is the -- not only the searching and
` 3 matching, but the evolution of what it is you're
` 4 searching and having that be adaptive. And he --
` 5 and so I was interested in generalizing certain
` 6 ideas for adaptive database growth, if you will, in
` 7 the lossless context to images and video and had
` 8 initially recruited him to do experiments.
` 9 Oh, and I'm thinking of another one too.
`10 So there -- actually, this one particular topic has
`11 a number of graduate students, but in any case, you
`12 wanted an example, and I think my initial answer was
`13 that when you're an academic, you normally work in
`14 conjunction with graduate students doing Ph.D.
`15 theses, and the work can involve contributions from
`16 everyone. It could be the design was mine and the
`17 implementation was theirs. It could be improvements
`18 made by them, improvements made by me, all different
`19 kinds of combinations over the years with many
`20 different