SRI INTERNATIONAL

Patent Trials and Appeals Board

May 5, 2020

2019004226 (P.T.A.B. May. 5, 2020)

UNITED STATES PATENT AND TRADEMARK OFFICE
UNITED STATES DEPARTMENT OF COMMERCE
United States Patent and Trademark Office
Address: COMMISSIONER FOR PATENTS
P.O. Box 1450
Alexandria, Virginia 22313-1450
www.uspto.gov
APPLICATION NO. FILING DATE FIRST NAMED INVENTOR ATTORNEY DOCKET NO. CONFIRMATION NO.
13/916,346 06/12/2013 Jeremy Epstein SRI6167-2 1361
14824 7590 05/05/2020
Moser Taboada/ SRI International
1030 Broad Street
Suite 203
Shrewsbury, NJ 07702
EXAMINER
GREGG, MARY M
ART UNIT PAPER NUMBER
3697
NOTIFICATION DATE DELIVERY MODE
05/05/2020 ELECTRONIC
Please find below and/or attached an Office communication concerning this application or proceeding.
The time period for reply, if any, is set in the attached communication.
Notice of the Office communication was sent electronically on above-indicated "Notification Date" to the
following e-mail address(es):
docketing@mtiplaw.com
llinardakis@mtiplaw.com
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
BEFORE THE PATENT TRIAL AND APPEAL BOARD
Ex parte JEREMY EPSTEIN
Appeal 2019-004226
Application 13/916,346
Technology Center 3600
Before ERIC S. FRAHM, BETH Z. SHAW, and
SCOTT B. HOWARD, Administrative Patent Judges.
SHAW, Administrative Patent Judge.
DECISION ON APPEAL
STATEMENT OF THE CASE
Pursuant to 35 U.S.C. § 134(a), Appellant1 appeals from the
Examiner’s decision to reject claims 2, 3, and 7–23. See Final Act. 1. We
have jurisdiction under 35 U.S.C. § 6(b).
We AFFIRM.

1 We use the word Appellant to refer to “applicant” as defined in 37 C.F.R.
§ 1.42. Appellant identifies the real party in interest as SRI
INTERNATIONAL. Appeal Br. 3.
Appeal 2019-004226
Application 13/916,346

2
CLAIMED SUBJECT MATTER
The claims are directed to a method for performing transaction
authorization to an online system from an untrusted computer system. Claim
2, reproduced below, is illustrative of the claimed subject matter:
2. A computer-implemented method for allowing a user of an
online system to authorize transactions from untrusted computer
systems, the method comprising executing on a processor the
steps of, with the online system:
receiving a transaction request of the user, the transaction
request to accomplish an online transaction, the transaction
request comprising financial information including at least an
amount of the transaction, and in response to the transaction
request:
presenting a series of tests to the user, each of the tests
comprising text that is human intelligible but not intelligible by
computers and at least a portion of the financial information of
the transaction request including at least the amount of the
transaction;
creating, using a microphone, a recording of the user’s
voice speaking one of the tests of the series of tests and the
portion of the financial information;
performing speaker verification, using a speaker model, on
the recording to determine if the recording matches a known
sample of the user’s voice;
performing automated speech recognition, using an
acoustic model and a language model, on the recording to
determine if the spoken financial information matches the
financial information of the transaction request and if the spoken
one of the tests matches one of the presented series of tests; and
rejecting the transaction request to prevent fraudulent
transactions if at least one of a) the one of the tests spoken by the
user does not match one of the presented series of tests, b) the
recording does not match the known sample of the user’s voice
and c) the recognized spoken financial information does not
match the financial information of the transaction request.
Appeal 2019-004226
Application 13/916,346

3
REFERENCES
The prior art relied upon by the Examiner is:
Name Reference Date
Bird US 2007/0179885 A1 Aug. 2, 2007
Berstis US 2007/0186184 A1 Aug. 9, 2007
Lee US 2008/0140420 A1 June 12, 2008
Weller US 2010/0114776 A1 May 6, 2010
Hulten US 2011/0314537 A1 Dec. 22, 2011
Gross US 8,494,854 B2 July 23, 2013

REJECTION
In the Answer, the Examiner withdrew the rejection of the claims
under 35 U.S.C. § 101. Ans. 3.
Claims 2, 3, 7–10, 12, 13, 15, 16, 18–21, 23, and 24 are rejected under
35 U.S.C. § 103 as being unpatentable over Gross and Lee. Final Act. 48,
68, 69.
Claim 11 is rejected under 35 U.S.C. § 103 as being unpatentable over
Gross, Lee, and Bird. Final Act. 65.
Claim 14 is rejected under 35 U.S.C. § 103 as being unpatentable over
Gross, Lee, and Berstis. Final Act. 66.
Claim 17 is rejected under 35 U.S.C. § 103 as being unpatentable over
Gross, Lee, Berstis, and Hulten. Final Act. 68.
Claim 22 is rejected under 35 U.S.C. § 103 as being unpatentable over
Gross, Lee, and Weller. Final Act. 78.

Appeal 2019-004226
Application 13/916,346

4
OPINION
Section 103 Rejections
Appellant argues the Examiner erred in rejecting the claims. We have
considered and reviewed the Examiner’s rejections in light of Appellant’s
arguments that the Examiner has erred. We are not persuaded of error for the
following reasons, and we agree with and adopt the Examiner’s findings and
conclusions in the Final Rejection (Final Act. 48–80) and Answer (Ans. 3–
20).
Appellant argues that Gross or Lee, or their combination, fail to teach
or suggest various disputed limitations of independent claim 2, as discussed
in more detail below. See Appeal Br. 18–29.
Presenting a Series of Tests
Appellant’s claim includes receiving a transaction request, including
an amount of the transaction, presenting a series of tests to a user that
include the amount of the transaction, and creating a recording of the user
speaking the amount of the transaction. Appeal Br. 19.
Appellant argues that Lee and Gross do not teach or suggest
“presenting a series of tests to the user, each of the tests comprising . . . at
least a portion of the financial information of the transaction request
including at least the amount of the transaction,” as recited in claim 2.
Appeal Br. 26. In particular, Appellant argues that Lee merely teaches that
words spoken by a user of the remote payment device to verify an identity
can include a currency name and transaction amount. Id.
Appellant’s arguments regarding Lee’s individual shortcomings in
this regard are unavailing, because the Examiner does not rely solely on Lee
for teaching the disputed series of tests limitation, but rather the cited prior
Appeal 2019-004226
Application 13/916,346

5
art collectively. Final Act. 49–54. Therefore, Appellant’s arguments do not
show nonobviousness where, as here, the rejection is based on the cited
references’ collective teachings. In particular, the Examiner explains, and we
agree, that Gross presents a series of tests to the user, each test including text
that is human-intelligible but not intelligible by computers and at least a
portion of the financial information. Final Act. 49 (citing Gross, 3:50–67;
4:1–59; 5:45–61; 6:10–23). The Examiner additionally finds, and we agree,
that Lee teaches the transaction request including at least the amount of the
transaction. Final Act. 52, 53 (citing Lee ¶¶ 7–10, 38, 52–57, Figs. 1, 7, 8).
Appellant acknowledges that Lee teaches that a first voice “consists of
several words of which can be the currency name and the transaction
amount.” Appeal Br. 26.
Accordingly, Appellant’s arguments with respect to this disputed
limitation are not persuasive.

Performing Speaker Verification on the Recording
Appellant argues Gross fails to teach “performing speaker
verification, using a speaker model, on the recording to determine if the
recording matches a known sample of the user’s voice,” where the recoding
includes “text that is human intelligible but not intelligible by computers and
at least a portion of the financial information of the transaction request
including at least the amount of the transaction,” as recited in claim 2.
Appeal Br. 21. “That is, in the Appellant[’s] independent claim 2, the
Appellant[] claims performing speaker verification, using a speaker model,
on the recording.” Id. “As such, in accordance with the Appellant[’s]
claims, the recording includes text that is human intelligible but not
Appeal 2019-004226
Application 13/916,346

6
intelligible by computers and at least a portion of the financial information
of the transaction request including at least the amount of the
transaction.” Id. Appellant argues that in contrast to Appellant’s claim,
“Gross only teaches using speech waveforms of human utterances models to
train a device to differentiate between human and machine utterances.” Id. at
22, 23.
Appellant acknowledges that “Gross teaches that speaker verification
systems are used to make a reliable determination if the identity of a person
matches a prior recorded voice print.” Id.
Moreover, the Examiner finds, and we agree, that Lee teaches the
disputed speaker verification because Lee teaches:
Before authorizing the charge, the user inputs a first voice to the
speech-recognition piece by speaking, and then the speech-recognition
piece will recognize if input speech is the amount to be paid, and
verify the similarity of the first voice with the predetermined voice. If
the first voice is recognized and matched to the transaction
amount, and falls within the recognition zone, the charge will be
allowed, which means the user passes the authorization of the
remote payment system.

Ans. 6 (quoting Lee ¶ 38). The Examiner finds, and we agree, that Lee
teaches matching the first voice to a previous voice spoken to the remote
payment system (Lee ¶ 43 (“voice is exactly the voice the owner spoke to
the remote payment system 1 before”)), “describing an amount of money”
(Lee ¶ 7), and “recording a speech” (Lee ¶ 6). Ans. 6. Accordingly, Lee
teaches or suggests, “performing speaker verification, using a speaker
model, on the recording to determine if the recording matches a known
sample of the user’s voice,” as recited in independent claim 2.
Appeal 2019-004226
Application 13/916,346

7
Additionally, the Examiner relies on Lee, not Gross alone, to teach the
transaction request “including at least the amount of the transaction.” Final
Act. 53 (citing Lee ¶¶ 7–10, 38, 52–57). We agree with the Examiner that
Lee teaches fraud prevention when a “transaction amount” is “inputted by a
first voice.” Lee ¶ 57. Therefore, Appellant’s arguments do not show
nonobviousness where, as here, the rejection is based on the cited
references’ collective teachings.

Performing Automated Speech Recognition Using an Acoustic Model and a
Language Model
Appellant also argues that Gross fails to teach or suggest “performing
automated speech recognition, using an acoustic model and a language
model, on the recording to determine if the spoken financial information
matches the financial information of the transaction request and if the
spoken one of the tests matches one of the presented series of tests,” recited
in independent claim 2. Appeal Br. 24. In particular, Appellant argues that
Gross does not teach automated speech recognition using a “language
model” on the recording. Id.; Reply Br. 26.
Regardless of the general contentions and imputed intended meanings
articulated by Appellant in the Appeal Brief, “[i]t is the claims that measure
the invention.” See SRI Int’l v. Matsushita Elec. Corp. of Am., 775 F.2d
1107, 1121 (Fed. Cir. 1985) (en banc) (citations omitted).
Though understanding the claim language may be aided by the
explanations contained in the written description, it is important
not to import into a claim limitations that are not a part of the
claim. For example, a particular embodiment appearing in the
written description may not be read into a claim when the claim
language is broader than the embodiment.
Appeal 2019-004226
Application 13/916,346

8
SuperGuide Corp. v. DirecTV Enters, Inc., 358 F.3d 870, 875 (Fed. Cir.
2004) (citing Electro Med. Sys. S.A. v. Cooper Life Sci., Inc., 34 F.3d 1048,
1054 (Fed. Cir. 1994)). “[A]lthough the specification often describes very
specific embodiments of the invention, [the Federal Circuit has] repeatedly
warned against confining the claims to those embodiments.” Phillips v.
AWH Corp., 415 F.3d 1303, 1323 (Fed. Cir. 2005) (en banc) (citations
omitted).
The Examiner finds, and we agree, that Gross and Lee teach this
disputed element, i.e., the claimed “performing automated speech
recognition, using an acoustic model and a language model, on the recording
to determine if the spoken financial information matches the financial
information of the transaction request and if the spoken one of the tests
matches one of the presented series of tests.” Claim 2; Ans. 11–12, 15–16.
The Examiner explains how Gross teaches that a spoken test can be a verbal
challenge test and how Lee teaches that recording the spoken test can
include financial information, including matching the transaction amount.
Id.; Lee ¶ 38.
Appellant provides insufficient evidence proving that the
Specification or claims limit “language model” in a way that, under a broad
but reasonable interpretation, is not encompassed by Gross and Lee’s
teachings on speech recognition. See Ans. 15–16. Moreover, we note that
Gross describes acoustic models (Hidden Markov Model) and that
computing systems “can be selected based on a language spoken by the
entity.” Gross, 2:26–28, 4:19–24; Final Act. 50.
Accordingly, Appellant’s arguments with respect to this disputed
limitation are not persuasive.
Appeal 2019-004226
Application 13/916,346

9

Rejecting the Transaction Request to Prevent Fraudulent Transactions
Appellant argues Gross and Lee fail to teach or suggest, “rejecting the
transaction request to prevent fraudulent transactions if at least one of a) the
one of the tests spoken by the user does not match one of the presented
series of tests, b) the recording does not match the known sample of the
user’s voice and c) the recognized spoken financial information does not
match the financial information of the transaction request,” as recited in
claim 2. Appeal Br. 27.
First, we note that claim 2 requires only “at least one” of conditions a,
b, or c, to be satisfied to reject the transaction request. The Examiner finds
that Lee teaches this disputed element. Ans. 18. We agree with the Examiner
that Lee teaches fraud prevention when a “transaction amount” is “inputted
by a first voice.” Lee ¶ 57. This is because Lee explains that a remote
payment system “will identify the first voice to confirm if the user is the real
owner.” Id. Thus, here, Lee at least suggests rejecting the transaction if the
voice is not that of the real owner.
Moreover, Lee goes on to describe another fraud prevention
mechanism, where the system detects if the voice inputted is “too similar” to
a predetermined voice, then the system asks the user to input a second voice
to be verified. Id. The phrase “too similar” “means the result falls in one of
three zones based on the predetermined voice.” Id. “The identification
results can roughly be divided into three types: within the recognition zone,
below the recognition zone and over the recognition zone. If the
identification results falls below the recognition zone, the remote payment
system will recognize the inputted voice as a wrong voice and implement the
Appeal 2019-004226
Application 13/916,346

10
step of ending directly.” Id. (emphasis added). Thus, this portion of Lee also
teaches “rejecting the transaction request to prevent fraudulent transactions
if . . . b) the recording does not match the known sample of the user’s
voice,” because Lee will end (reject) the transaction if the inputted voice is
the “wrong voice.” Lee ¶ 57.
In the Reply Brief, Appellant argues that Lee does not teach that the
remote payment system uses a test, recording, or financial information.
Reply Br. 9. We are not persuaded by this argument because Lee uses a
recording (element b of the disputed limitation) as explained above. Lee ¶
57.
Moreover, the Examiner does not rely on Lee alone to teach the series
of test element of the claim, but rather, as explained above, the Examiner
relies on the combination of Lee and Gross to teach a series of tests. Thus,
Appellant’s arguments regarding Lee’s individual shortcomings in this
regard (Reply Br. 9) are unavailing, for the Examiner does not rely solely on
Lee for teaching the test, but rather the cited prior art collectively.
Therefore, Appellant’s arguments do not show nonobviousness where, as
here, the rejection is based on the cited references’ collective teachings.

Conclusion of Obviousness
Finally, Appellant does not argue the Examiner’s rationale in
combining the references, and we discern no error therein. As the Examiner
finds, both references relate to voice testing technology with the motivation
of preventing fraud to provide greater security, and each discloses a variety
of voice technology alternatives that a person of ordinary skill would have
recognized as being applicable to improve security in the other. Final Act.
Appeal 2019-004226
Application 13/916,346

11
55; see Innovention Toys LLC v. MGA Entertainment Inc., 637 F.3d 1314,
1321 (Fed. Cir. 2011); see id. at 1322–23 (finding no error in combining
references that share the “same purpose,” “goal,” or “objective”). Moreover,
in an obviousness analysis, prior art references must be “considered together
with the knowledge of one of ordinary skill in the pertinent art,” In re
Paulsen, 30 F.3d 1475, 1480 (Fed. Cir. 1994), and we “take account of the
inferences and creative steps that a person of ordinary skill in the art would
employ.” KSR Int’l Co. v. Teleflex Inc., 550 U.S. 398, 418 (2007). On this
record, we find (and Appellant does not dispute) the Examiner has
articulated ample “reasoning with some rational underpinning to support the
legal conclusion of obviousness.” KSR, 550 U.S. at 418; Ans.
Accordingly, we sustain the rejection of independent claim 2 under
§ 103. For the same reasons, we sustain the rejection of independent claim
20, for which Appellant presents the same or essentially the same arguments.
See Appeal Br. 29, 32, 33. For the same reasons, we sustain the rejection of
claims 3, 7–10, 12, 13, 15, 16, 18, 19, 21, 23, and 24, for which Appellant
presents no additional arguments. Id.
We also sustain the various rejections of dependent claims 11, 14, 17,
and 22 because Appellant presents the same or essentially the same
arguments for these claims as for claim 2. See Appeal Br. 29–31.
CONCLUSION
The Examiner’s rejection is affirmed.
We affirm the rejection of claims 2, 3, 7–10, 12, 13, 15, 16, 18–21,
23, and 24 under 35 U.S.C. § 103 as being unpatentable over Gross and Lee.
We affirm the rejection of claim 11 under 35 U.S.C. § 103 as being
unpatentable over Gross, Lee, and Bird.
Appeal 2019-004226
Application 13/916,346

12
We affirm the rejection of claim 14 under 35 U.S.C. § 103 as being
unpatentable over Gross, Lee, and Berstis.
We affirm the rejection of claim 17 under 35 U.S.C. § 103 as being
unpatentable over Gross, Lee, Berstis, and Hulten.
We affirm the rejection of claim 22 under 35 U.S.C. § 103 as being
unpatentable over Gross, Lee, and Weller.

DECISION SUMMARY
Claims
Rejected
35 U.S.C. § Reference(s)/Basis Affirmed Reversed
2, 3, 7–10,
12, 13, 15,
16, 18–21,
23, 24
103 Gross, Lee 2, 3, 7–10,
12, 13, 15,
16, 18–21,
23, 24

11 103 Gross, Lee, Bird 11
14 103 Gross, Lee, Berstis 14
17 103 Gross, Lee, Berstis,
Hulten
17
22 103 Gross, Lee, Weller 22
Overall
Outcome:
2, 3, 7–23

TIME PERIOD FOR RESPONSE
No time period for taking any subsequent action in connection with
this appeal may be extended under 37 C.F.R. § 1.136(a). See 37 C.F.R.
§ 1.136(a)(1)(iv).
AFFIRMED