`
`·2· · · · · · · ----------------------------
`
`·3· · · · BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`·4· · · · · · · ----------------------------
`
`·5· · · · · · · · · · · ·APPLE INC.
`
`·6· · · · · · · · · · · ·Petitioner
`
`·7· · · · · · · · · · · · · ·v.
`
`·8· · · · · · ·ANDREA ELECTRONICS CORPORATION
`
`·9· · · · · · · · · · · Patent owner
`
`10· · · · · · · -----------------------------
`
`11· · · · · · · · ·Case No. IPR2017-00626
`
`12· · · · · · · · ·U.S. Patent 6,363,345
`
`13· · · · · · · · ·Case No. IPR2017-00627
`
`14· · · · · · · · · U.S. Patent 5,550,924
`
`15
`
`16· · · · DEPOSITION OF BERTRAND M. HOCHWALD, Ph.D.
`
`17· · · · · · · · · · Washington, D.C.
`
`18· · · · · · · · · · October 12, 2017
`
`19· · · · · · · · · · · · 9:38 a.m.
`
`20
`
`21
`
`22· Job No. BO-143617
`
`23· Pages 1 - 193
`
`24· Reported by:· Michele E. Eddy, RPR, CRR, CLR
`
`Patent Owner
`Andrea Electronics Corp.
`EXHIBIT 2005
`IPR2017-00627
`
`IPR Page 1 of 221
`
`
`
`·1· · · · · · · Deposition of BERTRAND M. HOCHWALD, Ph.D.,
`
`·2· held at the offices of:
`
`·3· · · · · · SIDLEY AUSTIN LLP
`· · · · · · · 1501 K Street, Northwest
`·4· · · · · · Washington, D.C.· 20005
`
`·5· · · · · · (202) 736-8000
`
`·6
`
`·7· · · · · · Pursuant to Notice, before Michele E.
`
`·8· Eddy, Registered Professional Reporter, Certified
`
`·9· Realtime Reporter, and Notary public in and for
`
`10· the District of Columbia.
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`IPR Page 2 of 221
`
`
`
`·1· · · · · · · · · A P P E A R A N C E S
`
`·2
`
`·3· ON BEHALF OF THE PETITIONER:
`
`·4· · · · · · THOMAS A. BROUGHAN, III, ESQUIRE
`
`·5· · · · · · Sidley Austin LLP
`
`·6· · · · · · 1501 K Street, Northwest
`
`·7· · · · · · Washington, D.C.· 20005
`
`·8· · · · · · Telephone:· (202) 736-8000
`
`·9· · · · · · tbroughan@sidley.com
`
`10· · · · · · - AND -
`
`11· · · · · · STEVEN S. BAIK, ESQUIRE
`
`12· · · · · · Sidley Austin LLP
`
`13· · · · · · Building One, Suite 100
`
`14· · · · · · 1001 Page Mill Road
`
`15· · · · · · Palo Alto, California· 94304
`
`16· · · · · · Telephone:· (650) 565-7000
`
`17· · · · · · sbaik@sidley.com
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`IPR Page 3 of 221
`
`
`
`·1· ATTENDANCE, Continued
`
`·2
`
`·3· ON BEHALF OF THE PATENT OWNER:
`
`·4· · · · · · BRADLEY T. LENNIE, ESQUIRE
`
`·5· · · · · · SEAN T. GLOTH, ESQUIRE, LLP
`
`·6· · · · · · Pepper Hamilton LLP
`
`·7· · · · · · 600 Fourteenth Street, Northwest
`
`·8· · · · · · Washington, D.C.· 20005
`
`·9· · · · · · Telephone:· (202) 220-1200
`
`10· · · · · · lennieb@pepperlaw.com
`
`11· · · · · · gloths@pepperlaw.com
`
`12· · · · · · - AND -
`
`13· · · · · · STEVEN R. PEDERSEN, ESQUIRE
`
`14· · · · · · The Pedersen Firm
`
`15· · · · · · 13 West River Road, Suite D
`
`16· · · · · · Rumson, New Jersey· 07760
`
`17· · · · · · Telephone:· (312) 342-8600
`
`18· · · · · · Steve@Pedersenfirm.com
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`IPR Page 4 of 221
`
`
`
`·1· · · · · · · · ·EXAMINATION INDEX
`
`·2· · · · · · · · · · · · · · · · · · · · · · · · · · PAGE
`
`·3· EXAMINATION BY MR. LENNIE· · · · · · · · · · · · · · 7
`
`·4· EXAMINATION BY MR. BROUGHAN· · · · · · · · · · · · 186
`
`·5
`
`·6
`
`·7
`
`·8
`
`·9· · · · · · · · · · ·E X H I B I T S
`
`10· · · · · · · (Attached to the Transcript)
`
`11· DEPOSITION EXHIBIT· · · · · · · · · · · · · · · · PAGE
`
`12· Exhibit 1· Petition for Inter Partes Review· · · · ·11
`
`13· · · · · · ·of U.S. Patent No. 6,363,345
`
`14· Exhibit 2· Figure 2:· Flowchart of the SNR· · · · · 30
`
`15· · · · · · ·estimation algorithm
`
`16· Exhibit 3· Petition for Inter Partes Review· · · · 142
`
`17· · · · · · ·of U.S. Patent No. 5,550,924
`
`18
`
`19· · · · · · PREVIOUSLY MARKED AND REFERRED TO
`
`20· PATENT OWNER· · · · · · · · · · · · · · · · · · · PAGE
`
`21· Exhibit 1001· U.S. Patent No. 6,363,345· · · · · · · 8
`
`22· Exhibit 1003· Declaration of Bertrand Hochwald· · · 12
`
`23· · · · · · · · Regarding U.S. Patent No. 6,363,345
`
`24
`
`IPR Page 5 of 221
`
`
`
`·1· · · · · · · · ·EXHIBIT INDEX CONTINUED
`
`·2
`
`·3· PATENT OWNER· · · · · · · · · · · · · · · · · · · PAGE
`
`·4· Exhibit 1004· Declaration of Bertrand Hochwald· · ·141
`
`·5· · · · · · · · Regarding U.S. Patent No. 5,550,924
`
`·6· Exhibit 1005· Noise Estimation Techniques for· · · ·15
`
`·7· · · · · · · · Robust Speech Recognition by H.G.
`
`·8· · · · · · · · Hirsch and C. Ehrlicher
`
`·9· Exhibit 1006· Abstract by Rainer Martin titled· · · 15
`
`10· · · · · · · · "An Efficient Algorithm to
`
`11· · · · · · · · Estimate the Instantaneous SNR
`
`12· · · · · · · · of Speech Signals"
`
`13· Exhibit 1010· U.S. Patent No. 5,550,924· · · · · · 142
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`23
`
`24
`
`IPR Page 6 of 221
`
`
`
`·1· · · · · · · · · P R O C E E D I N G S
`
`·2· · · · · · · · · · Washington, D.C.
`
`·3· · · · · · · ·October 12, 2017, 9:38 A.M.
`
`·4· · · · · · · · · · · · · - - -
`
`·5· · · · · · · · · · · · · - - -
`
`·6· · · · · · · BERTRAND M. HOCHWALD, Ph.D.,
`
`·7· having been duly sworn, testified as follows:
`
`·8· · · ·EXAMINATION BY COUNSEL FOR THE PATENT OWNER
`
`·9· BY MR. LENNIE:
`
`10· · · ·Q· · Good morning, Dr. Hochwald.
`
`11· · · ·A· · Good morning.
`
`12· · · ·Q· · Please state your full name for the
`
`13· record.
`
`14· · · ·A· · Bertrand Martin Hochwald.
`
`15· · · ·Q· · And where are you currently employed?
`
`16· · · ·A· · At the University of Notre Dame.
`
`17· · · ·Q· · And what is your position there?
`
`18· · · ·A· · I'm on the faculty there at the
`
`19· department of electrical engineering.
`
`20· · · ·Q· · Could you explain how you first got
`
`21· involved with this case.
`
`22· · · · · · MR. BROUGHAN:· I will caution the
`
`23· witness not to reveal the substance of
`
`24· communications with counsel.· Subject to that, you
`
`IPR Page 7 of 221
`
`
`
`·1· may answer.
`
`·2· · · ·A· · I'm not sure if I recall the exact
`
`·3· details of that.· It would have been sometime last
`
`·4· year, receiving a call from a representative of
`
`·5· Sidley or -- I don't remember.
`
`·6· · · ·Q· · And did you perform a search for prior
`
`·7· art related to the '345 patent?
`
`·8· · · · · · MR. BROUGHAN:· Again, I'll object to the
`
`·9· extent that it calls for privileged or work
`
`10· product protected information.· I'll caution the
`
`11· witness not to reveal the substance of
`
`12· communications with counsel.· Subject to that, you
`
`13· may answer.
`
`14· · · ·A· · Yes.
`
`15· · · ·Q· · You did.· Okay.
`
`16· · · · · · MR. LENNIE:· Let's go ahead and mark the
`
`17· '345 patent first.· So we're just going to use the
`
`18· exhibit numbers from the IPR proceeding.· So this
`
`19· is going to be Exhibit 1001.
`
`20· · · · · · (Exhibit 1001 was previously marked and
`
`21· referred to.)
`
`22· · · · · · MR. LENNIE:· So we have here Exhibit
`
`23· 1001 from the 626 petition.· So when I'm referring
`
`24· to the 626 petition, that's the IPR proceeding
`
`IPR Page 8 of 221
`
`
`
`·1· with the last three digits 626.
`
`·2· BY MR. LENNIE:
`
`·3· · · ·Q· · We've marked the '345 patent.· You're
`
`·4· familiar with the '345 patent.· Correct?
`
`·5· · · ·A· · Yes.
`
`·6· · · ·Q· · And you said that you conducted a prior
`
`·7· art search regarding the '345 patent.· Is that
`
`·8· correct?
`
`·9· · · ·A· · Yes.
`
`10· · · ·Q· · Did you do that personally, or did
`
`11· someone else do it?
`
`12· · · · · · MR. BROUGHAN:· Again, I'll caution the
`
`13· witness not to reveal the substance of
`
`14· communications with counsel.
`
`15· · · ·A· · It would have been personally.
`
`16· · · ·Q· · Personally?
`
`17· · · ·A· · Yes.
`
`18· · · ·Q· · How did you go about doing that search?
`
`19· · · ·A· · I guess the usual way, library searches,
`
`20· Internet searches, database, patent database
`
`21· searches.
`
`22· · · ·Q· · And in connection with that prior
`
`23· search, did you identify the references that you
`
`24· relied on in the 626 petition?
`
`IPR Page 9 of 221
`
`
`
`·1· · · · · · MR. BROUGHAN:· I'll caution the witness
`
`·2· again not to reveal the substance -- not to reveal
`
`·3· the substance of communications with counsel.
`
`·4· Subject to that, you may answer.
`
`·5· · · ·A· · I don't recall what I recovered exactly
`
`·6· or what I found myself.
`
`·7· · · ·Q· · And did you first review the '345 patent
`
`·8· prior to conducting that search?
`
`·9· · · ·A· · Yes.
`
`10· · · ·Q· · So did you have the '345 patent claims
`
`11· in mind when you were preparing that search?
`
`12· · · ·A· · I guess the whole patent claims and
`
`13· specification.
`
`14· · · ·Q· · And are you familiar with
`
`15· Dr. Kyriakakis?
`
`16· · · · · · MR. BROUGHAN:· I'll caution the witness
`
`17· not to reveal the substance of communications with
`
`18· counsel.· Subject to that, you may answer.
`
`19· · · ·A· · I don't know him personally.
`
`20· · · ·Q· · Have you ever met Dr. Kyriakakis?
`
`21· · · ·A· · Not that I recall.
`
`22· · · ·Q· · Have you ever spoken on the phone with
`
`23· Dr. Kyriakakis?
`
`24· · · ·A· · No.
`
`IPR Page 10 of 221
`
`
`
`·1· · · ·Q· · Let's go ahead and mark the petition and
`
`·2· your declaration.
`
`·3· · · · · · Let's mark this as Hochwald Exhibit 1.
`
`·4· · · · · · (Exhibit 1 was marked for identification
`
`·5· and attached to the deposition transcript.)
`
`·6· · · · · · MR. BROUGHAN:· At some point later --
`
`·7· well, never mind.
`
`·8· · · · · · So just with the exhibit numbering, I
`
`·9· didn't know if you were going to file this as an
`
`10· exhibit.· But I assume you're not since it's
`
`11· already in the record.
`
`12· BY MR. LENNIE:
`
`13· · · ·Q· · Sir, we've marked as Exhibit 1 a copy of
`
`14· Apple's petition on the '345 patent that was given
`
`15· the number IPR2017-626.· Are you familiar with
`
`16· this document?
`
`17· · · ·A· · Petition, no, I don't think so.· No.
`
`18· · · ·Q· · You haven't reviewed a copy of the
`
`19· petition that we marked as Exhibit 1?
`
`20· · · ·A· · Sorry, is this -- is this my
`
`21· declaration?
`
`22· · · ·Q· · This is a copy of the petition.
`
`23· · · ·A· · No, I don't -- I don't recall the
`
`24· petition.
`
`IPR Page 11 of 221
`
`
`
`·1· · · · · · (Exhibit 1003 was previously marked and
`
`·2· referred to.)
`
`·3· · · · · · MR. LENNIE:· I'll hand the witness a
`
`·4· copy of Exhibit 1003, which is a copy of
`
`·5· Dr. Hochwald's declaration submitted with the 626
`
`·6· petition.
`
`·7· BY MR. LENNIE:
`
`·8· · · ·Q· · Are you familiar with what we've marked
`
`·9· as Exhibit 1003?
`
`10· · · ·A· · Yes.
`
`11· · · ·Q· · In connection with preparing -- well,
`
`12· let me back up.
`
`13· · · · · · Could you explain at a high level how
`
`14· you prepared this declaration.
`
`15· · · · · · MR. BROUGHAN:· I'll object to the extent
`
`16· it calls for privileged or work product protected
`
`17· information.· I'll caution you not to reveal the
`
`18· substance of communications with counsel.· Subject
`
`19· to that, you may answer.
`
`20· · · ·A· · In what way?· Looking up prior art, what
`
`21· you're asking about before?
`
`22· · · ·Q· · Yes.· Let's start at the early stages.
`
`23· So the declaration identifies various references
`
`24· that you assert render the claims of the '345
`
`IPR Page 12 of 221
`
`
`
`·1· patent invalid.· Is that correct?
`
`·2· · · ·A· · Yes.
`
`·3· · · ·Q· · And how did you decide to identify the
`
`·4· references that are relied upon in your
`
`·5· declaration?
`
`·6· · · · · · MR. BROUGHAN:· Again, I'll caution the
`
`·7· witness not to reveal the substance of
`
`·8· communications with counsel.· Subject to that, you
`
`·9· may answer.
`
`10· · · ·A· · I'm not sure I know exactly how to
`
`11· answer that.· It's prior art that was found either
`
`12· by me or with the assistance of counsel.
`
`13· · · ·Q· · You said that you had prepared or
`
`14· conducted a prior art search in connection with
`
`15· your work in preparing this declaration.· Correct?
`
`16· · · ·A· · Yes.
`
`17· · · ·Q· · And did you identify additional prior
`
`18· art that's not discussed and relied upon in your
`
`19· declaration?
`
`20· · · ·A· · Possibly, but I included in the
`
`21· declaration what was needed to support the
`
`22· argument.
`
`23· · · ·Q· · And what was the process to determine
`
`24· which particular references and which particular
`
`IPR Page 13 of 221
`
`
`
`·1· combinations you would rely on in forming your
`
`·2· opinions regarding the '345 patent claims?
`
`·3· · · · · · MR. BROUGHAN:· I'll object to the extent
`
`·4· it calls for privileged or work product protected
`
`·5· information.· I'll caution the witness not to
`
`·6· reveal the substance of communications with
`
`·7· counsel.· Subject to that, you may answer.
`
`·8· · · ·A· · Sorry, I lost the thread of the
`
`·9· question.
`
`10· · · ·Q· · How did you decide which prior art
`
`11· references to rely on in connection with the
`
`12· opinions in your declaration that we marked as
`
`13· Exhibit 1003?
`
`14· · · · · · MR. BROUGHAN:· Same objections, same
`
`15· caution.
`
`16· · · ·A· · We, in reviewing the patent and its
`
`17· claims, looked at what material we thought was --
`
`18· or that I thought was going to be most relevant
`
`19· for anticipating the patent.
`
`20· · · ·Q· · So two of the prior art references that
`
`21· you rely on in your declaration are the Hirsch
`
`22· reference and the Martin reference.· Correct?
`
`23· · · ·A· · What -- can I see them?
`
`24· · · ·Q· · Yes.
`
`IPR Page 14 of 221
`
`
`
`·1· · · · · · (Exhibit 1005 was previously marked and
`
`·2· referred to.)
`
`·3· · · ·Q· · Exhibit 1005 is a copy of an article
`
`·4· titled, "Noise Estimation Techniques for Robust
`
`·5· Speech Recognition," authored by H.G. Hirsch and
`
`·6· C. Ehrlicher.· Are you familiar with Exhibit 1005?
`
`·7· · · ·A· · Yes.
`
`·8· · · ·Q· · I'll refer to 1005 as "the Hirsch
`
`·9· reference" or "Hirsch."· Is that fair?
`
`10· · · ·A· · Yes.
`
`11· · · · · · (Exhibit 1006 was previously marked and
`
`12· referred to.)
`
`13· · · ·Q· · I'm going to hand you what's been marked
`
`14· as Exhibit 1006.· Exhibit 1006 to the 626 petition
`
`15· is a copy of an article entitled, "An Efficient
`
`16· Algorithm to Estimate the Instantaneous SNR of
`
`17· Speech Signals," by Rainer Martin.
`
`18· · · · · · Are you familiar with what has been
`
`19· marked as Exhibit 1006?
`
`20· · · ·A· · Yes.
`
`21· · · ·Q· · You rely on -- I'll refer to Exhibit
`
`22· 1006 as the Martin reference.· Is that fair?
`
`23· · · ·A· · Yes.
`
`24· · · ·Q· · You rely on both the Hirsch reference
`
`IPR Page 15 of 221
`
`
`
`·1· and the Martin reference in your analysis of the
`
`·2· validity of the '345 patent claims.· Is that
`
`·3· correct?
`
`·4· · · ·A· · Yes.
`
`·5· · · ·Q· · And your -- is it your opinion that the
`
`·6· Hirsch reference anticipates claims 4 to 11 of the
`
`·7· '345 patent?
`
`·8· · · ·A· · Well, in my declaration I say Hirsch --
`
`·9· I don't specifically call out -- is there a
`
`10· specific spot in my declaration that you're
`
`11· referring to?
`
`12· · · ·Q· · Why don't you take a look at the Table
`
`13· of Contents.· It's on Roman numeral iii.
`
`14· · · ·A· · Background?
`
`15· · · ·Q· · Analysis of the Prior Art and the '345
`
`16· Claims --
`
`17· · · ·A· · Yes.
`
`18· · · ·Q· · -- Section VII.
`
`19· · · ·A· · Okay.
`
`20· · · ·Q· · So do you have an opinion as to whether
`
`21· or not the Hirsch reference anticipates claims 4
`
`22· to 11 of the '345 patent?
`
`23· · · ·A· · I have a -- so if I look at my
`
`24· declaration, Hirsch -- I describe Hirsch in the
`
`IPR Page 16 of 221
`
`
`
`·1· context of -- on page 44 and elements of claims 1
`
`·2· through 3, 12, 13, 21, 23, and 38.· And then
`
`·3· Martin 4 through 11, 29 through 42.· So I -- I'm
`
`·4· not sure how to answer your question.
`
`·5· · · ·Q· · Do you understand the difference between
`
`·6· anticipation and obviousness?
`
`·7· · · ·A· · Yes.
`
`·8· · · ·Q· · With respect to invalidity analysis?
`
`·9· · · ·A· · Yes.
`
`10· · · ·Q· · So let me ask again.
`
`11· · · · · · Is it your opinion that the Hirsch
`
`12· reference anticipates claims 4 to 11 of the '345
`
`13· patent?
`
`14· · · ·A· · Well, to the extent that my declaration
`
`15· combines -- let me just see.· (Document review.)
`
`16· · · · · · This one -- I have to look at the
`
`17· patent.· (Document review.)
`
`18· · · · · · So Hirsch -- you're asking the context
`
`19· of Hirsch by itself?
`
`20· · · ·Q· · The question was whether you have an
`
`21· opinion as to whether the Hirsch reference
`
`22· anticipates claims 4 to 11 of the '345 patent.
`
`23· · · ·A· · So Hirsch by itself.
`
`24· · · ·Q· · That's correct.
`
`IPR Page 17 of 221
`
`
`
`·1· · · ·A· · I don't seem -- I don't have an opinion
`
`·2· on that listed in my declaration.
`
`·3· · · ·Q· · What did you do to prepare for today's
`
`·4· deposition?
`
`·5· · · · · · MR. BROUGHAN:· I'll object to the extent
`
`·6· it calls for privileged or work product protected
`
`·7· information.· I caution the witness not to reveal
`
`·8· the substance of communications with counsel.
`
`·9· Subject to that, you may answer.
`
`10· · · ·A· · I came here on Tuesday and spent the
`
`11· last two days preparing, discussing the material.
`
`12· · · ·Q· · And did you review a copy of your
`
`13· declaration that we've marked as Exhibit 1003 in
`
`14· connection with your preparation for today's
`
`15· deposition?
`
`16· · · ·A· · Yes.
`
`17· · · ·Q· · And in your declaration, you do not
`
`18· contend that the Hirsch reference anticipates
`
`19· claims 4 to 11 of the '345 patent.· Isn't that
`
`20· correct?
`
`21· · · ·A· · Not by itself, no.
`
`22· · · ·Q· · Do you have an opinion as to whether
`
`23· claims 4 to 11 of the '345 patent are invalid?
`
`24· · · ·A· · Again, to the extent that the
`
`IPR Page 18 of 221
`
`
`
`·1· declaration says so, yes.
`
`·2· · · ·Q· · That's what I'm asking you.· Does the
`
`·3· declaration make the contention that the claims 4
`
`·4· to 11 of the '345 patent are invalid?
`
`·5· · · ·A· · Yes.
`
`·6· · · ·Q· · And what is your basis for your opinion
`
`·7· that the claims 4 to 11 of the '345 patent are
`
`·8· invalid?
`
`·9· · · ·A· · I say on page 54 that Martin describes
`
`10· the elements of the claims 4 through 11.
`
`11· · · ·Q· · Is it your opinion that the Martin
`
`12· reference anticipates claims 4 to 11 of the '345
`
`13· patent?
`
`14· · · ·A· · (Document review.)
`
`15· · · · · · I'm sorry, Martin by itself?
`
`16· · · ·Q· · Yes.
`
`17· · · ·A· · Yes.
`
`18· · · ·Q· · It's your opinion that the Martin
`
`19· reference anticipates claims 4 to 11 of the '345
`
`20· patent?
`
`21· · · ·A· · Yes.
`
`22· · · ·Q· · Can you point me to where in your
`
`23· declaration you offer that opinion or that
`
`24· conclusion.
`
`IPR Page 19 of 221
`
`
`
`·1· · · ·A· · Starting on page -- or paragraph 133.
`
`·2· · · ·Q· · Do you understand that claim 4 depends
`
`·3· from claim 1 of the '345 patent?
`
`·4· · · ·A· · Yes.
`
`·5· · · ·Q· · Do you understand that claim 4 includes
`
`·6· all the elements of claim 1?
`
`·7· · · ·A· · Yes.
`
`·8· · · ·Q· · Do you have an opinion as to whether the
`
`·9· Martin reference anticipates claim 1 of the '345
`
`10· patent?
`
`11· · · ·A· · Again, by itself, I have no opinion of
`
`12· that in my declaration.
`
`13· · · ·Q· · So your declaration -- in your
`
`14· declaration you don't contend that the Martin
`
`15· reference anticipates claim 1 of the '345 patent?
`
`16· · · ·A· · That's correct.
`
`17· · · ·Q· · So if Martin does not anticipate claim 1
`
`18· of the '345 patent, Martin cannot anticipate claim
`
`19· 4 of the '345 patent.· Is that correct?
`
`20· · · ·A· · I didn't contend that -- there's two
`
`21· things there.· I didn't contend that Martin
`
`22· doesn't anticipate claim 1, but claim 4 is more
`
`23· restrictive than claim 1.· So I'm not sure I
`
`24· understand the logic.
`
`IPR Page 20 of 221
`
`
`
`·1· · · ·Q· · Does your declaration rely on a
`
`·2· combination of Hirsch and Martin with respect to
`
`·3· the alleged invalidity of claims 4 to 11 of the
`
`·4· '345 patent?
`
`·5· · · ·A· · So my declaration talks about Martin by
`
`·6· itself in the context of claims 4 through 11.
`
`·7· · · ·Q· · How did you determine that the elements
`
`·8· of claim 4 were shown in the prior art?
`
`·9· · · ·A· · Well, I would read claim 4 of the '345,
`
`10· the specification, and then examine the prior art
`
`11· to see if I saw substantially the same thing.
`
`12· · · ·Q· · Did you start your analysis of the
`
`13· alleged invalidity of the '345 patents -- '345
`
`14· patent with claim 1?
`
`15· · · ·A· · Yes.
`
`16· · · ·Q· · And it's your opinion that Hirsch
`
`17· discloses all of the elements of claim 1.· Is that
`
`18· correct?
`
`19· · · ·A· · Yes.
`
`20· · · ·Q· · But it's your -- is it also your opinion
`
`21· that Hirsch does not disclose the elements of
`
`22· claim 4?
`
`23· · · ·A· · No.· My declaration just uses Hirsch in
`
`24· the context of claims 1 through 3, 12, 13, 21, 23,
`
`IPR Page 21 of 221
`
`
`
`·1· and 38.
`
`·2· · · ·Q· · Do you have an opinion as to whether
`
`·3· Hirsch discloses the elements of claim 4 of the
`
`·4· '345 patent?
`
`·5· · · ·A· · No, I don't.· I don't have an opinion on
`
`·6· that.
`
`·7· · · ·Q· · Let's take a look at the Martin
`
`·8· reference, which we've marked as Exhibit 1006.
`
`·9· · · · · · Can you describe at a high level the
`
`10· algorithm that the Martin article describes.
`
`11· · · ·A· · So Martin -- Martin is interested in the
`
`12· context of speech signals and speech processing to
`
`13· estimate the noise spectrum of the -- of a speech
`
`14· signal in the context of reducing the noise.
`
`15· · · ·Q· · Does the Martin reference employ windows
`
`16· and subwindows in its algorithm?
`
`17· · · ·A· · Yes.· Martin has a parameter W called
`
`18· the window, and I believe -- I believe there's a
`
`19· subwindow, although I can't find it right at the
`
`20· moment.
`
`21· · · ·Q· · So what does W represent in the Martin
`
`22· algorithm?
`
`23· · · ·A· · On page 1094, "decomposed into W windows
`
`24· of length M," so window.
`
`IPR Page 22 of 221
`
`
`
`·1· · · ·Q· · So does W represent the number of
`
`·2· subwindows?
`
`·3· · · ·A· · I'm not sure I can answer that.· I don't
`
`·4· know the context of the subwindows.
`
`·5· · · ·Q· · Let's look at page 1094 of Martin.
`
`·6· · · · · · Under the bottom of the left-hand column
`
`·7· under "Noise power estimation" -- do you see that
`
`·8· section?
`
`·9· · · ·A· · Yes.
`
`10· · · ·Q· · -- the second sentence of that section
`
`11· states, "For reasons of computational complexity
`
`12· and delay, the data window of length L is
`
`13· decomposed into W windows of length M such that M
`
`14· times W equals L."
`
`15· · · · · · Do you see that?
`
`16· · · ·A· · Yes.
`
`17· · · ·Q· · Is it fair to say that the Martin
`
`18· reference employs windows of length L and multiple
`
`19· subwindows that will correspond to the variable W?
`
`20· · · ·A· · Well, Martin calls them windows so I
`
`21· don't know the context of the subwindows.
`
`22· · · ·Q· · It says that "the data window of length
`
`23· L is decomposed into W windows of length M."
`
`24· · · · · · Do you see that?
`
`IPR Page 23 of 221
`
`
`
`·1· · · ·A· · Yes.
`
`·2· · · ·Q· · Does that tell you that the window
`
`·3· length L is comprised of subwindows of length W?
`
`·4· · · ·A· · I guess we can call them subwindows, but
`
`·5· that's not what Martin uses.
`
`·6· · · ·Q· · It's true that Martin does describe L
`
`·7· being divided into further divisions of W windows.
`
`·8· Correct?
`
`·9· · · ·A· · Uh-hmm.· Yes.· Sorry.
`
`10· · · ·Q· · And you need to answer yes or no.
`
`11· · · ·A· · Right.· Sorry.
`
`12· · · ·Q· · So for the purposes of the deposition,
`
`13· I'm going to refer to the window length L as a
`
`14· window and the W windows that comprise that L
`
`15· window as subwindows.· Is that fair?
`
`16· · · ·A· · Okay.
`
`17· · · ·Q· · Now, in your analysis of the Martin
`
`18· reference, do you assume W to be equal to a
`
`19· particular number?
`
`20· · · ·A· · (Document review.)
`
`21· · · · · · Yes.· On page 58 I assume -- I say, set
`
`22· W equal to 1.
`
`23· · · ·Q· · So you said page 58 of your declaration?
`
`24· · · ·A· · Yes.
`
`IPR Page 24 of 221
`
`
`
`·1· · · ·Q· · And does that assumption that W equals 1
`
`·2· apply to all of your analysis with respect to the
`
`·3· Martin reference?
`
`·4· · · ·A· · Sure --
`
`·5· · · · · · MR. BROUGHAN:· Object to form.
`
`·6· · · · · · THE WITNESS:· Sorry.
`
`·7· · · · · · MR. BROUGHAN:· I object to form.
`
`·8· · · ·Q· · If your attorney objects, you still have
`
`·9· to answer the question.
`
`10· · · ·A· · Okay.· So W equals 1 applies to at least
`
`11· some of my analysis.
`
`12· · · ·Q· · Which analysis is that?
`
`13· · · ·A· · Well, the part where I say W equals 1
`
`14· starting at paragraph 136.
`
`15· · · ·Q· · If W equals 2 or more, do your opinions
`
`16· with respect to the Martin reference still hold
`
`17· true?
`
`18· · · ·A· · Yes.· I think so, yes.
`
`19· · · ·Q· · Which sections of your declaration, if
`
`20· any, analyze the Martin reference where W is equal
`
`21· to 2 or more?
`
`22· · · ·A· · Yes, in section 136 I say W equals 1 for
`
`23· purposes of the illustration, but I don't say
`
`24· anything specific about W in other portions.· So I
`
`IPR Page 25 of 221
`
`
`
`·1· don't believe I made any assumptions about W in
`
`·2· the parts where I don't say anything.
`
`·3· · · ·Q· · So if we look at paragraph 137 on page
`
`·4· 58 of your declaration --
`
`·5· · · ·A· · Yes.
`
`·6· · · ·Q· · -- the first sentence there indicates
`
`·7· that you're assuming -- making an assumption about
`
`·8· Martin's algorithm.· Do you see that?
`
`·9· · · ·A· · I see "Assume that Martin's algorithm,"
`
`10· yes.
`
`11· · · ·Q· · In the rest of that sentence, you're
`
`12· making the assumption that W equals 1.
`
`13· · · ·A· · Yes.
`
`14· · · ·Q· · There you use the term "subwindows."
`
`15· Correct?
`
`16· · · ·A· · Yes.
`
`17· · · ·Q· · So you're familiar with the term
`
`18· "subwindows"?
`
`19· · · ·A· · Yes.· I use that term here to
`
`20· distinguish it from the windows.
`
`21· · · ·Q· · That's your term.· You're describing the
`
`22· number of windows within window length L as
`
`23· subwindows.· Is that correct?
`
`24· · · ·A· · Correct.
`
`IPR Page 26 of 221
`
`
`
`·1· · · ·Q· · Then in paragraph 139 you're also noting
`
`·2· that either of the lower boxes circled in blue
`
`·3· Pn(i) equals PMmin because W equals 1.· Correct?
`
`·4· · · ·A· · Yes.
`
`·5· · · ·Q· · Is that assumption that W equals 1
`
`·6· applicable throughout the rest of your
`
`·7· declaration?
`
`·8· · · · · · MR. BROUGHAN:· Object to the form.
`
`·9· · · ·A· · Only where I use W equals 1 explicitly
`
`10· can I answer that, and only in those locations do
`
`11· I assume W equals 1.· I make no assumptions about
`
`12· W in other parts.
`
`13· · · ·Q· · All right.· I want you to assume a
`
`14· couple assumptions with respect to how the Martin
`
`15· algorithm works.· So if you assume that L is equal
`
`16· to .156 seconds.· Do you have that?
`
`17· · · ·A· · Hold on one second.· L is an integer, I
`
`18· think.· It's number of samples.
`
`19· · · ·Q· · L is the length of the overall window.
`
`20· · · ·A· · Sorry, I don't understand the context.
`
`21· · · ·Q· · It's on page 1094.· If you look at page
`
`22· 1094 of the Martin reference, it's the bottom of
`
`23· the left-hand column going into the top of the
`
`24· right-hand column.· We read this sentence earlier.
`
`IPR Page 27 of 221
`
`
`
`·1· "For reasons of computational complexity and
`
`·2· delay, the data window of length L is decomposed
`
`·3· into W windows of length M such that M times W
`
`·4· equals L."
`
`·5· · · · · · Do you see that?
`
`·6· · · ·A· · Yes.
`
`·7· · · ·Q· · So L is the length of the overall
`
`·8· window.· Correct?
`
`·9· · · ·A· · Yes.
`
`10· · · ·Q· · And W is the number of subwindows.
`
`11· Correct?
`
`12· · · ·A· · Okay, yes.
`
`13· · · ·Q· · I want you to assume that L in this
`
`14· hypothetical is equal to .156 seconds.· Okay?
`
`15· · · ·A· · Again, the L -- I'm not sure I follow
`
`16· that L -- L should be an integer for this to work,
`
`17· I think.
`
`18· · · ·Q· · If you go back to the Martin reference,
`
`19· to page 1094, it discloses in the very next
`
`20· sentence that we read earlier that "For a sampling
`
`21· rate of f8 equals 8 kHz, typical window parameters
`
`22· are M equals 1250 and W equals 4, thus equals L
`
`23· equals 5000 corresponding to a time window of
`
`24· 0.625 s."
`
`IPR Page 28 of 221
`
`
`
`·1· · · · · · Do you see that?
`
`·2· · · ·A· · Yes.
`
`·3· · · ·Q· · So is that stating that L corresponds to
`
`·4· a time window of .625 seconds?
`
`·5· · · ·A· · Yes.
`
`·6· · · ·Q· · Okay.· So I would like you to assume
`
`·7· that L corresponds to a time window of .156
`
`·8· seconds.· Does that make sense?
`
`·9· · · ·A· · I'm not sure it does because there are
`
`10· other parameters here, such as the sampling rate
`
`11· and the window parameters, like the M and the W,
`
`12· to give me an L value that's an integer.· So if
`
`13· you ask me for, I think, a time window of .156
`
`14· seconds, I don't know what L that corresponds to.
`
`15· · · · · · MR. LENNIE:· Let's go off the record and
`
`16· take a quick break.
`
`17· · · · · · · ·(A brief recess was taken.)
`
`18· · · · · · MR. LENNIE:· We can go back on the
`
`19· record.
`
`20· BY MR. LENNIE:
`
`21· · · ·Q· · Dr. Hochwald, at the break did you have
`
`22· an opportunity to talk to your attorneys at all?
`
`23· · · ·A· · No.
`
`24· · · ·Q· · So the Martin reference, as we
`
`IPR Page 29 of 221
`
`
`
`·1· discussed, discloses an example where L is equal
`
`·2· to 5000 that corresponds to a time window of .625
`
`·3· seconds.· Is that correct?
`
`·4· · · ·A· · Yes.
`
`·5· · · ·Q· · It also discusses W equal 4, which would
`
`·6· mean there are four subwindows.· Is that correct?
`
`·7· · · ·A· · Yes.
`
`·8· · · ·Q· · On page 1094 of the Martin reference,
`
`·9· figure 2 identifies a flowchart of the SNR
`
`10· estimation algorithm.· Is that correct?
`
`11· · · ·A· · Yes, that's the title.
`
`12· · · ·Q· · Are you familiar with figure 2?
`
`13· · · ·A· · Yes.
`
`14· · · ·Q· · Figure 2 is a little bit difficult to
`
`15· read in the copy of the reference that we have so
`
`16· I'm going to mark as Exhibit 2 a reproduced copy
`
`17· of figure 2.
`
`18· · · · · · (Exhibit 2 was marked for identification
`
`19· and attached to the deposition transcript.)
`
`20· BY MR. LENNIE:
`
`21· · · ·Q· · Just take a minute just to confirm to
`
`22· your own satisfaction that what we've marked as
`
`23· Exhibit 2 is the same diagram, flowchart that's
`
`24· shown in figure 2 of the Martin reference.
`
`IPR Page 30 of 221
`
`
`
`·1· · · ·A· · Yes, it looks like it.
`
`·2· · · ·Q· · And you're familiar with figure 2 and
`
`·3· how figure 2 describes the flowchart of Martin's
`
`·4· algorithm.· Right?
`
`·5· · · ·A· · Yes.
`
`·6· · · ·Q· · All right.· I want to direct your
`
`·7· attention to the first diamond block in figure 2.
`
`·8· · · ·A· · Okay.
`
`·9· · · · · · MR. BROUGHAN:· Object to form.
`
`10· · · ·Q· · And this is the block that includes
`
`11· Px(i) is less than PMmin.
`
`12· · · · · · Do you see that?
`
`13· · · ·A· · Px with the bar on it.
`
`14· · · ·Q· · Yes.
`
`15· · · ·A· · Yes.
`
`16· · · ·Q· · What does Px(i) represent in this block?
`
`17· · · · · · MR. BROUGHAN:· Object to form,
`
`18· foundation.
`
`19· · · ·A· · Martin characterizes it as the smoothed
`
`20· short time power estimate.
`
`21· · · ·Q· · And what does PMmin represent?
`
`22· · · ·A· · I'm not sure of the wording used by
`
`23· Martin.· It appears -- it appears to be the power
`
`24· minimum -- power minimum is the best I can do from
`
`IPR Page 31 of 221
`
`
`
`·1· the reference.
`
`·2· · · ·Q· · Your declaration relies extensively on
`
`·3· Martin's disclosure of PMmin in your invalidity
`
`·4· analysis.· Isn't that correct?
`
`·5· · · ·A· · Yes.
`
`·6· · · ·Q· · What does PMmin represent in this first
`
`·7· diamond block of figure 2?
`
`·8· · · ·A· · So PMmin is a variable, and Martin
`
`·9· attaches a name to it and uses it in the context
`
`10· of an algorithm.· What it means in the particular
`
`11· portion of the algorithm is subject to the context
`
`12· because the algorithm has a beginning and end and
`
`13· middle.
`
`14· · · ·Q· · What does PMmin represent in this first
`
`15· diamond where the diamond states Px(i) is less
`
`16· than PMmin?
`
`17· · · ·A· · What it means in any particular portion
`
`18· is, I think, very hard to say and, in particular,
`
`19· in that first diamond.
`
`20· · · ·Q· · What is the comparison being done in the
`
`21· diamond block Px(i) is less than PMmin?
`
`22· · · ·A· · It's checking to see if this condition
`
`23· Px(i) bar less than PMmin is satisfied.
`
`24· · · ·Q· · And what does Px(i) represent?
`
`IPR Page 32 of 221
`
`
`
`·1· · · ·A· · Px bar (i) represents the smoothed short
`
`·2· time power estimate, according to Martin.
`
`·3· · · ·Q· · So what happens if PMmin is greater than
`
`·4· Px(i) in this first top diamond in figure 2?
`
`·5· · · ·A· · So if the condition is met, then PMmin
`
`·6· is set to Px bar (i).
`
`·7· · · ·Q· · What happens if PMmin is less than
`
`·8· Px(i)?
`
`·9· · · ·A· · Then it passes through to the next
`
`10· block, next diamond.
`
`11· · · ·Q· · Assume for purposes of these next
`
`12· questions that W equals 4.· So we have four
`
`13· subwindows.· Correct?
`
`14· · · ·A· · Okay.
`
`15· · · ·Q· · And then further assume that L equals
`
`16· 5000 and corresponds to a time window of .625
`
`17· seconds.· Okay?
`
`18· · · ·A· · Okay.
`
`19· · · ·Q· · So with those assumptions, in this
`
`20· diamond block that we've been discussing, how many
`
`21· times will PMmin be replaced with Px(i)?
`
`22· · · · · · MR. BROUGHAN:· Object to form.
`
`23· · · ·A· · I don't think I can answer that without
`
`24· more information.· I haven't considered that in my
`
`IPR Page 33 of 221
`
`
`
`·1· declaration.
`
`·2· · · ·Q· · What information -- what additional
`
`·3· information