Ex Parte Pearce et al

Board of Patent Appeals and Interferences

May 6, 2009

09830306 (B.P.A.I. May. 6, 2009)

UNITED STATES PATENT AND TRADEMARK OFFICE
____________

BEFORE THE BOARD OF PATENT APPEALS
AND INTERFERENCES
____________

Ex parte DAVID JOHN BENJAMIN PEARCE and
JON ALASTAIR GIBBS
____________

Appeal 2008-5318
Application 09/830,306
Technology Center 2600
____________

Decided:1 May 7, 2009
____________

Before KENNETH W. HAIRSTON, CARLA M. KRIVAK, and
KARL D. EASTHOM, Administrative Patent Judges.

HAIRSTON, Administrative Patent Judge.

DECISION ON APPEAL

1 The two-month time period for filing an appeal or commencing a civil
action, as recited in 37 CFR § 1.403, begins to run from the decided dated
shown on this page of the decision. The time period does not run from the
Mail Date (paper delivery) or Notification Date (electronic delivery).
Appeal 2009-5318
Application 09/830,306

Appellants appeal under 35 U.S.C. § 134 from a final rejection of
claims 1, 3, 4, 6, 7, 11 to 13, 15, 16, 18, 19, and 23 to 26. We have
jurisdiction under 35 U.S.C. § 6(b).
We will sustain the obviousness rejections.
Appellants’ invention is concerned with mitigating transmission errors
in a distributed speech recognition method and system by replacing one or
more speech recognition parameters in an identified group of vectors that
have a transmission error with corresponding copies of one or more
corresponding speech recognition parameters from a different vector (Figs. 1
and 2; Spec. 9 to 11).
Claim 1 is representative of the claims on appeal, and it reads as
follows:
1. A method of mitigating errors in a distributed speech recognition
process, the distributed speech recognition process being one in which
speech recognition parameters are arranged in vectors, each vector
corresponding to a particular sampling time-frame, and said speech
recognition parameters are received at a second location having been
transmitted from a first location;
the method comprising the steps of:
identifying a group comprising one or more of said vectors which
have undergone a transmission error; and
replacing one or more speech recognition parameters in the identified
group of vectors, wherein said one or more speech recognition parameters in
said identified group of vectors are replaced by respective replacement
parameters corresponding to copies of one or more corresponding speech
recognition parameters from a different vector, corresponding to a different
2
Appeal 2009-5318
Application 09/830,306

particular sampling time frame, received without error after said identified
group of vectors.
The prior art relied upon by the Examiner in rejecting the claims on
appeal is:
Ozawa US 5,305,332 Apr. 19, 1994
Jeon US 5,673,363 Sep. 30, 1997
Yeldener US 5,774,837 Jun. 30, 1998
Jacobs US 5,956,683 Sep. 21, 1999
The Examiner rejected claims 1, 3, 13, 15, 25, and 26 under 35 U.S.C.
§ 103(a) based upon the teachings of Jacobs and Jeon.
The Examiner rejected claims 4 and 16 under 35 U.S.C. § 103(a)
based upon the teachings of Jacobs, Jeon, and Ozawa.
The Examiner rejected claims 6, 7, 11, 12, 18, 19, 23, and 24 under 35
U.S.C. § 103(a) based upon the teachings of Jacobs, Jeon, and Yeldener.
Appellants and the Examiner agree that Jacobs does not provide any
type of transmission error correction (Br. 6; Ans. 4). Appellants argue inter
alia (Br. 6 to 8) that the applied references do not teach replacing one or
more speech recognition parameters or the entire vector that has undergone a
transmission error with copies of one or more corresponding speech
recognition parameters.
ISSUE
Have Appellants demonstrated that the Examiner erred by finding that
the reference to Jeon teaches replacing one or more speech recognition
parameters or the entire vector that has undergone a transmission error with
copies of one or more corresponding speech recognition parameters?

3
Appeal 2009-5318
Application 09/830,306

FINDINGS OF FACT
1. The transmission error mitigation performed by Appellants replaces
speech recognition parameters in a vector (e.g., vector 133) with
corresponding speech recognition parameters from a different vector (e.g.,
vector 134, see Fig. 1).
2. Jacobs describes a distributed speech recognition process in which
speech recognition parameters are arranged in vectors, and each vector
corresponds to a particular sampling time frame (Fig. 2; Abstract; col. 1, ll.
22 to 29).
3. In the preferred embodiment of Jeon, an error in a segment of a
frame is concealed by taking an adjacent segment from another frame and
multiplying the adjacent segment by a weight value to obtain a coefficient
for the erroneous segment (Fig. 6A; col. 4, l. 62 to col. 5, l. 18).
4. In the admitted prior art of Jeon, a frame with erroneous segment(s)
is replaced in toto with the contents of a buffered error-free frame (Figs. 3
and 4A; col. 1, ll. 5 to 12; col. 1, l. 59 to col. 2, l. 31).
PRINCIPLES OF LAW
“[I]f a technique has been used to improve one device, and a person of
ordinary skill in the art would recognize that it would improve similar
devices in the same way, using the technique is obvious unless its actual
application is beyond his or her skill.” KSR Int’l Co. v. Teleflex Inc., 550
U.S. 398, 417 (2007).
ANALYSIS
Claims 1, 13, 25, and 26
Inasmuch as the preferred embodiment in Jeon conceals an error in a
segment of a frame by taking an adjacent segment from another frame and
4
Appeal 2009-5318
Application 09/830,306

multiplying the adjacent segment by a weight value to obtain a coefficient
for the erroneous segment, we agree with Appellants’ argument that such an
error concealment technique does not teach replacing one or more speech
recognition parameters or the entire vector that has undergone a transmission
error with copies of one or more corresponding speech recognition
parameters. As seen in Figure 6A of the preferred embodiment in Jeon
(FF3), the adjacent segment Sm from frame F1 is not a corresponding copy of
erroneous segment S1 from frame F2. On the other hand, the admitted prior
art in Figures 3 and 4A of Jeon (FF4) describes the replacement in toto of a
frame F2 with erroneous segment(s) with another frame from buffer 40.
Thus, the admitted prior art portion of Jeon teaches replacing one or more
speech recognition parameters or the entire vector that has undergone a
transmission error with copies of one or more corresponding speech
recognition parameters as set forth in claims 1, 13, 25, and 26 on appeal. In
the absence of other arguments by Appellants, we find that it would have
been manifestly obvious to one of ordinary skill in the art to use the error
mitigation technique described by the admitted prior art in Jeon to improve
the distributed speech recognition process described by Jacobs (FF2). See
KSR, 550 U.S. at 417.
Claims 3 and 15
Appellants’ arguments (Br. 8) for claims 3 and 15 mirror the
arguments presented for claims 1, 13, 25, and 26. Accordingly, Appellants’
arguments do not demonstrate the nonobviousness of the claimed subject
matter set forth in these claims.
5
Appeal 2009-5318
Application 09/830,306

Claims 4 and 16
Appellants have chosen to let these claims stand or fall with claims 3
and 15 (Br. 9). As indicated supra, Appellants’ arguments were not
convincing of the nonobviousness of the claimed subject matter set forth in
claims 3 and 15.
Claims 6, 7, 18, and 19
Appellants have chosen to let these claims stand or fall with claims 1
and 13 (Br. 9). As indicated supra, Appellants’ argument were not
convincing of the nonobviousness of the claimed subject matter set forth in
claims 1 and 13.
Claims 11, 12, 23, and 24
Appellants’ arguments (Br. 9) fail to address the Examiner’s
contention that it would have been obvious to the skilled artisan to apply the
teachings of Yeldener to the teachings of Jacobs and Jeon. Accordingly,
Appellants’ arguments do not overcome the Examiner’s prima facie case of
obviousness.
CONCLUSION OF LAW
Appellants have not demonstrated that the Examiner erred by finding
that the applied references teach replacing one or more speech recognition
parameters or the entire vector that has undergone a transmission error with
copies of one or more corresponding speech recognition parameters.
ORDER
The Examiner’s obviousness rejections of claims 1, 3, 4, 6, 7, 11 to
13, 15, 16, 18, 19, and 23 to 26 are affirmed.
6
Appeal 2009-5318
Application 09/830,306

No time period for taking any subsequent action in connection with
this appeal may be extended under 37 CFR § 1.136(a)(1)(iv).

AFFIRMED

ELD

JONATHAN P MEYER
MOTOROLA INC
1303 EAST ALGONQUIN ROAD
SCHAUMBURG, IL 60196
7