Ex Parte Singh et al

Patent Trials and Appeals Board

Jan 28, 2019

15243635 - (D) (P.T.A.B. Jan. 28, 2019)

UNITED STA TES p A TENT AND TRADEMARK OFFICE
APPLICATION NO. FILING DATE
15/243,635 08/22/2016
63649 7590 01/30/2019
DISNEY ENTERPRISES, INC.
C/0 FARJAMI & FARJAMI LLP
26522 LA ALAMEDA A VENUE, SUITE 360
MISSION VIEJO, CA 92691
FIRST NAMED INVENTOR
Rita Singh
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.
0260492 9385
EXAMINER
HANG,VUB
ART UNIT PAPER NUMBER
2672
NOTIFICATION DATE DELIVERY MODE
01/30/2019 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@farj ami. com
farjamidocketing@yahoo.com
ffarj ami @farj ami. com
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
BEFORE THE PATENT TRIAL AND APPEAL BOARD
Ex parte RITA SINGH and JILL FAIN LEHMAN 1
Appeal2018-005675
Application 15/243,635
Technology Center 2600
Before CARLA M. KRIVAK, HUNG H. BUI, and JON M. JURGOV AN,
Administrative Patent Judges.
KRIVAK, Administrative Patent Judge.
DECISION ON APPEAL
Appellants appeal under 35 U.S.C. § 134(a) from the Examiner's
Final Rejection of claims 1-20, which are all the claims pending in the
application. We have jurisdiction under 35 U.S.C. § 6(b).
We affirm-in-part.
1 Appellants identify the real party in interest as Disney Enterprises, Inc.
(see App. Br. 2).
Appeal2018-005675
Application 15/243,635
STATEMENT OF THE CASE
Appellants' invention is directed to a method and system "for
estimating age of a speaker based on speech" by extracting "a plurality of
formant-based feature vectors from each phoneme in the digitized speech
based on at least one of a formant position, a formant bandwidth, and a
formant dispersion," and comparing the formant-based feature vectors with
age determinant formant-based feature vectors (Title ( capitalization altered);
Abstract).
Claims 1 and 9 are independent. Independent claim 1, reproduced
below, is exemplary of the subject matter on appeal.
1. A system comprising:
a microphone configured to receive an input speech from
an individual;
an analog-to-digital (AID) converter configured to
convert the input speech from an analog form to a digital form
and generate a digitized speech;
a memory storing an executable code and an age
estimation database including a plurality of age determinant
formant-based feature vectors;
a hardware processor executing the executable code to:
receive the digitized speech from the AID
converter;
identify a plurality of boundaries between a
plurality of phonemes in the digitized speech;
extract a plurality of formant-based feature vectors
from one or more phonemes of the plurality of phonemes
delineated by the plurality of boundaries, based on at
least one of a formant position, a formant bandwidth, and
a formant dispersion, wherein the formant dispersion is a
geometric mean of the formant spacing;
compare the plurality of formant-based feature
vectors with the age determinant formant-based feature
vectors of the age estimation database;
2
Appeal2018-005675
Application 15/243,635
estimate the age of the individual when the
comparison finds a match in the age estimation database;
and
communicate an age-appropriate response to the
individual based on the estimated age of the individual.
REJECTIONS and REFERENCES
The Examiner rejected claims 1, 3, 4, 6-9, 11, 12, and 14--20 under
35 U.S.C. § 103 based upon the teachings of Ju (US 2008/0298562 Al,
published Dec. 4, 2008) and Zhao (US 2005/0228664 Al, published Oct. 13,
2005).
The Examiner rejected claims 2, 5, 10, and 13 under 35 U.S.C. § 103
based upon the teachings of Ju, Zhao, and Zigel (US 2015/0351663 Al,
published Dec. 10, 2015).
ANALYSIS
Claims 1-16, 18, and 20
With respect to claim 1, Appellants contend "there is no disclosure,
teaching or suggestion in Zhao about a formant position," where (i)
"[f]ormant position may be the peak frequency of a formant" and (ii)
"'formant' means 'the spectral peaks of the sound spectrum !P(f)I' or 'a
concentration of acoustic energy around a particular frequency in the speech
wave"' (App. Br. 9, 12 (citing Spec. 14:5-10); Reply Br. 2). Appellants
further argue Zhao does not teach or suggest "extracting formant-based
feature vectors from phonemes delineated by the boundaries is based on a
formant position" as recited in claim 1 (Reply Br. 3--4). We do not agree.
We agree with and adopt the Examiner's findings as our own.
Particularly, we agree with the Examiner each ofZhao's (2N+ l)*m-
3
Appeal2018-005675
Application 15/243,635
dimension super vectors 406, including acoustic vectors, teaches aformant-
based feature vector extracted from phonemes ( 401, 404 in Zhao' s speech
waveform 400) delineated by a boundary (402), as required by claim 1 (see
Zhao ,r,r 30, 33-35, 46, Fig. 4; Ans. 3; Final Act. 4).
We also agree with the Examiner that Zhao teaches that feature
vectors ( acoustic vectors in the super vector 406) are extracted from
phonemes based on a formant position, as recited in claim 1 (Ans. 3). As
the Examiner correctly finds, Appellants' Specification does not provide an
explicit and exclusive definition of the claimed formant position; rather, it
broadly describes "a 'formant position' ... may be the peak frequency of a
formant" (Ans. 3 (citing Spec. 14:5-10)). Additionally, Appellants'
Specification does not define the claimed feature vectors, but merely
describes that "feature vectors [ are extracted] from phonemes of digitized
speech 108 corresponding to one or more formant-based measurements such
as formant positions, formant bandwidth, and/or formant dispersion" (see
Spec. 7: 17-19 ( emphasis added)). Based on Appellants' Specification, the
Examiner has broadly interpreted the claimed "formant-based feature
vectors . . . [ extracted] ... based on ... a formant position" as
encompassing Zhao' s acoustic feature vectors extracted from a speech
waveform at locations where acoustic energy is concentrated (Ans. 3). The
Examiner's interpretation and findings are reasonable because Zhao teaches
"[t]he acoustic features can be any of the widely used features such as
MFCCs (Mel Frequency Cepstral Coefficients), LPCs (Linear Prediction
Coefficients), LSPs (Line Spectral Pair)/LSF (Line Spectral Frequencies),
etc." (see Zhao ,r 35 (emphasis added)). Thus, a skilled artisan, viewing
Zhao' s teaching, would recognize Zhao' s acoustic features may be based on
4
Appeal2018-005675
Application 15/243,635
spectral peaks in the speech waveform or acoustic energy concentrations at
particular locations/frequencies ( formant positions), suggesting the claimed
extraction of features based on (well known), formant frequencies. 2
In light of the broad terms recited in claim 1 and the arguments
presented, Appellants have failed to clearly distinguish the claimed
invention over the prior art relied on by the Examiner. Thus, we sustain the
Examiner's rejection of independent claims 1 and 9 argued together and
dependent claims 2-8, 10-16, 18, and 20, not separately argued (App. Br. 7,
12-13).
Claims 17 and 19
Dependent claims 17 and 19 recite "the extracting of the plurality of
formant-based feature vectors [recited in claim 1 and 9] is based on the
formant position, the formant bandwidth, and the formant dispersion." The
2 It was known to the skilled artisan at the time of Appellants' invention that
cepstral coefficients, line spectral frequencies (LSFs ), and line spectrum pair
frequencies (LSPs ), may be based on spectral formant peaks or acoustic
energy concentrations at particular frequencies-see, e.g., Jonathan Lareau,
Application of shifted delta cepstral features for GMM language
identification (Oct. 10, 2006), (unpublished thesis, Rochester Institute of
Tech.) ( on file with the Thesis/Dissertation Collections at RIT Scholar
Works) ("an approximation procedure is used to first find the formant
envelope before deriving the Cepstral coefficients"), p. 30,
https://scholarworks.rit.edu/cgi/viewcontent.cgi?article=1260&context=thes
es; K.K. Paliwal, On the Use of Line Spectral Frequency Parameters for
Speech Recognition, 2 Digital Signal Processing 80-87 (1992) ( describing
LSFs based on formant frequencies at page 81); and Robert W. Morris &
Mark A. Clements, Modification of Formants in the Line Spectrum Domain,
9 IEEE SIGNAL PROCESSING LETTERS, 19, 19-21 (2002) ("we propose
an algorithm that takes advantage of the nearly linear relationship between
the LSPs and formants").
5
Appeal2018-005675
Application 15/243,635
Examiner finds Zhao's Figure 4 and paragraph 35 teaches the limitations of
claim 17 (Final Act. 6).
Appellants argue Zhao does not teach or suggest extraction of
formant-based feature vectors based on all of the recited factors including
"the formant bandwidth, and the formant dispersion" (Reply Br. 5; App. Br.
12-13). We agree with Appellants. Although Zhao's extracted acoustic
features "can be any of the widely used features such as MFCCs (Mel
Frequency Cepstral Coefficients), LPCs (Linear Prediction Coefficients),
LSPs (Line Spectral Pair)/LSF (Line Spectral Frequencies)," Zhao does not
teach extracting acoustic features based on both (i) formant bandwidth and
(ii) formant dispersion that is a geometric mean of the formant spacing, as
recited in claim 1, from which claim 17 depends (see Zhao ,r 35). As the
Examiner has not identified sufficient evidence to support the rejection of
claim 17, we do not sustain the Examiner's rejection of dependent claims 17
and 19.
DECISION
The Examiner's decision rejecting claims 1-16, 18, and 20 under
35 U.S.C. § 103 is affirmed.
The Examiner's decision rejecting claims 17 and 19 under
35 U.S.C. § 103 is reversed.
No time period for taking any subsequent action in connection with
this appeal may be extended under 37 C.F.R. § 1.136(a)(l )(iv).
AFFIRMED-IN-PART
6