Ex Parte Arakawa et al

Patent Trial and Appeal Board

Aug 30, 2016

11489594 (P.T.A.B. Aug. 30, 2016)

UNITED STA TES p A TENT AND TRADEMARK OFFICE
APPLICATION NO. FILING DATE FIRST NAMED INVENTOR
111489,594 0712012006 Takayuki Arakawa
21254 7590 08/31/2016
MCGINN INTELLECTUAL PROPERTY LAW GROUP, PLLC
8321 OLD COURTHOUSE ROAD
SUITE 200
VIENNA, VA 22182-3817
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
ATTORNEY DOCKET NO. CONFIRMATION NO.
NEG-613US 1370
EXAMINER
ORTIZ SANCHEZ, MICHAEL
ART UNIT PAPER NUMBER
2658
MAILDATE DELIVERY MODE
08/31/2016 PAPER
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.
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
BEFORE THE PATENT TRIAL AND APPEAL BOARD
Ex parte TAKA YUKI ARAKAWA and MASANORI TSUJIKA WA
Appeal2015-004005
Application 11/489,594
Technology Center 2600
Before THU A. DANG, NORMAN H. BEAMER,
and MONICA S. ULLAGADDI, Administrative Patent Judges.
BEAMER, Administrative Patent Judge.
DECISION ON APPEAL
Appellants appeal under 35 U.S.C. § 134(a) the Examiner's Final
Rejection of claims 1, 3-20, 22-25, and 27-30. 1 Claims 2, 21, and 26 are
cancelled. We have jurisdiction over the pending rejected claims under
35 U.S.C. § 6(b).
We reverse.
1 Appellants identify NEC Corporation as the real party in interest. (App.
Br. 1.)
Appeal2015-004005
Application 11/489,594
THE INVENTION
Appellants' disclosed and claimed invention is directed to noise
suppression in speech recognition. (Spec. p.1.) Claim 1, reproduced below,
is illustrative of the claimed subject matter:
1. A noise suppression system, comprising:
a unit, as executed by a processor, for successively
acquiring an input signal in a spectrum domain;
a unit, as executed by said processor, for successively
estimating an instant noise value in the spectrum domain from
said input signal;
a unit, as executed by said processor, for deriving a
provisional estimate speech in the spectral domain from said
input signal and said instant noise value; and
a unit, as executed by said processor, for correcting said
provisional estimate speech using a reference pattern of speech
stored in a storage unit, said correcting using a distribution for
said reference pattern as comprising clean speech without a
noise contamination,
wherein, in said unit for deriving said provisional
estimate speech, said provisional estimate speech is derived by
suppressing a noise element in said input signal with said
instant noise value, and
wherein said unit for correcting said provisional estimate
speech includes:
a unit for transforming said provisional estimate
speech derived in the spectral domain into a feature
vector in a logarithmic domain or a cepstrum domain;
a unit for correcting said provisional estimate
speech, transformed into said feature vector, using a
reference pattern in a feature vector domain;
a unit for transforming said corrected provisional
estimate speech in the spectrum domain; and
2
Appeal2015-004005
Application 11/489,594
a unit for acquiring an estimate speech by second
suppressing, in the spectrum domain, a noise element in
said input signal.
REJECTIONS
The Examiner rejected claims 1, 3, 4, 10, 11, 18, 19, 20, 22, 23, 25,
27, 28, and 30 under 35 U.S.C. § 103(a) as being unpatentable over Deng et
al. (US 2003/0225577 Al, pub. Dec. 4, 2003) and Kanazawa (US
2003/0177007 Al, pub. Sept. 18, 2003). (Final Act. 3-9.)
The Examiner rejected claims 5, 6, and 12-14 under 35 U.S.C.
§ 103(a) as being unpatentable over Deng, Kanazawa, and Segura et al.,
"Model-based compensation of the additive noise for continuous speech
recognition. Experiments using AURORA II database and tasks"
(Eurospeech 2001). (Final Act. 9-12.)
The Examiner rejected claims 7-9, 15-17, 24, and 29 under 35 U.S.C.
§ 103(a) as being unpatentable over Deng; Kanazawa; and Handel (US
5,943,429, issued Aug. 24, 1999). (Final Act. 12-16.)
ISSUE ON APPEAL
Appellants' arguments in the Appeal Brief present the following
dispositive issue: 2
2 Rather than reiterate the arguments of Appellants and the findings of the
Examiner, we refer to the Appeal Brief (filed Aug. 27, 2014); the Reply
Brief (filed Feb. 9, 2015); the Final Office Action (mailed Jan. 14, 2014);
and the Examiner's Answer (mailed Dec. 12, 2014) for the respective
details.
3
Appeal2015-004005
Application 11/489,594
Whether the Examiner erred in finding the combination of Deng and
Kanazawa teaches or suggests the independent claim 1 limitation, "a unit for
acquiring an estimate speech by second suppressing, in the spectrum
domain, a noise element in said input signal" (hereafter, the "second
suppressing" limitation), and the similar limitation recited in independent
claims 20 and 25. (App. Br. 13-17.)
ANALYSIS
In finding Deng and Kanazawa teach or suggest the second
suppressing limitation at issue, the Examiner relies on the disclosure in Deng
of a noise reduced vector "determined in an iterative fashion." (Final Act. 4;
Deng i-f 92.) Appellants, referring to the claim element at issue as the "final
step of this final limitation," argue:
The final step of this final limitation clearly recites that a
second suppression of noise occurs in the spectrum domain for
the corrected provisional estimate ( \'l1hich \'l/as obtained using a
reference pattern in the feature vector domain). Neither
primary reference Deng nor secondary reference Kanazawa
even suggests this unique sequence of noise suppression.
(App. Br. 13.) In response, the Examiner does not further refer to the
Paragraph 92 of Deng, but relies on Kanazawa:
[P]aragraph 104 of Kanawaza [sic] states that speech
recognition accuracy can be improved by replacing extracted
target voice spectrum information with reference spectrum
information so taught in the claims new limitation of a unit for
transforming said corrected estimate speech in the spectrum
domain is taught by Kanazawa.
(Ans. 15; see Kanazawa i-f 104.) Appellants argue, with respect to
Kanazawa:
4
Appeal2015-004005
Application 11/489,594
[T]he Examiner's reliance upon this paragraph [0104] is clearly
misplaced, since there seems to be no connection between the
content of this paragraph relative to the changes that the
Examiner identifies as necessary to primary reference Deng to
arrive at the claimed invention. Rather, this paragraph merely
describes that reference spectrum information with which no
noise is mixed is more accurate when noise is relatively small.
(App. Br. 16; see also Reply Br. 3--4.)
We are persuaded the Examiner errs. The Examiner does not
sufficiently explain how the broad reference to iterative noise reduction in
Deng, or the discussion of reference spectrum information in Kanazawa,
whether taken alone or in combination, teach or suggest the second
suppressing limitation at issue. Although, as a general proposition, the order
of elements in a claim do not necessarily require performance or operation in
the order stated, the Examiner does not take issue with Appellants' argument
that the "second suppressing" element at issue must take place "in the
spectrum domain for the corrected provisional estimate (which was obtained
using a reference pattern in the feature vector domain)." (App. Br. 13.)
Based on the literal language of the claims, and on the disclosures in the
Specification, we are persuaded by Appellants' argument. (Spec. Fig. 5.)
See Interactive Gift Express, Inc. v. Compuserve Inc., 256 F.3d 1323, 1342-
43 (Fed. Cir. 2001);Altiris, Inc. v. Symantec Corp., 318F.3d1363, 1370
(Fed. Cir. 2003) (citing Interactive Gift at 13 43: order may be determined by
the "logic or grammar" of the claim, or an explicit or implicit requirement in
the specification.)
Therefore, on the record before us, we are constrained to find the
Examiner errs in rejecting independent claims 1, 20, and 25.
5
Appeal2015-004005
Application 11/489,594
CONCLUSIONS
For the reasons stated above, we do not sustain the obviousness
rejection of independent claims 1, 20, and 25 over Deng and Kanazawa. We
also do not sustain the obviousness rejections of claims 3, 4, 10, 11, 18, 19,
22, 23, 27, 28, and 30 over Deng and Kanazawa, of claims 5, 6, and 12-14
over Deng, Kanazawa, and Segura, and of claims 7-9, 15-17, 24, and 29
over Deng, Kanazawa, and Handel, which claims depend from claims 1, 20,
or 25.
DECISION
The Examiner's rejections of claims 1, 3-20, 22-25, and 27-30 are
reversed. 3
REVERSED
3 In the event of further prosecution, we leave to the Examiner to decide
whether to reject claims 25-30 as being directed to nonstatutory subject
matter under 35 U.S.C. § 101. Claims 25-30 recite, "A computer program
product for use on a computer ... tangibly embodying a set of machine-
readable instructions for causing the computer to execute .... " However,
Appellants' Specification does not define such computer program product to
exclude transitory media. Consequently, the claimed medium encompasses
transitory media, which is not patent eligible. See Ex parte Mewherter, 107
USPQ2d 1857 (PTAB 2013) (precedential).
6