Continental Automotive Systems, Inc.

Patent Trials and Appeals Board

Aug 19, 2021

2020004159 (P.T.A.B. Aug. 19, 2021)

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.
14/636,562 03/03/2015 Bijal Joshi 2013P04246US 3895
64588 7590 08/19/2021
Continental Automotive Systems, Inc.
Patents & Licenses
21440 W Lake Cook Road
Floor 4
Deer Park, IL 60010
EXAMINER
HANG, VU B
ART UNIT PAPER NUMBER
2672
NOTIFICATION DATE DELIVERY MODE
08/19/2021 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):
patent@continental-corporation.com
patentsus@continental-corporation.com
robert.smolik@continental-corporation.com
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
_______________

BEFORE THE PATENT TRIAL AND APPEAL BOARD
_______________

Ex parte BIJAL JOSHI
_______________

Appeal 2020-004159
Application 14/636,5621
Technology Center 2600
_______________

Before ERIC B. CHEN, HUNG H. BUI, and ADAM J. PYONIN,
Administrative Patent Judges.

BUI, Administrative Patent Judge.

DECISION ON APPEAL
Appellant seeks our review under 35 U.S.C. § 134(a) from the
Examiner’s final rejection of claims 1–13, all the pending claims. We have
jurisdiction under 35 U.S.C. § 6(b).
We REVERSE.2

1 Herein, Appellant refers to “applicant(s)” as defined in 37 C.F.R. § 1.42.
Appellant identifies the real party in interest as Continental Automotive
Systems, Inc. Appeal Br. 2.
2 Our Decision refers to Appellant’s Appeal Brief filed October 30, 2019
(“Appeal Br.”); Reply Brief filed May 6, 2020 (“Reply Br.”); Examiner’s
Answer mailed February 6, 2020 (“Ans.”); Final Office Action mailed
November 29, 2018 (“Final Act.”); and original Specification filed March 3,
2015 (“Spec.”).
Appeal 2020-004159
Application 14/636,562

2

STATEMENT OF THE CASE
Appellant’s claimed invention relates to “a method and apparatus
[shown in Figure 3] for improving or restoring voice or speech fidelity or
speech quality after noise accompanying the speech is suppressed.”
Spec. ¶ 1. Appellant’s Figure 3 (annotated) is reproduced below.

Figure 3 shows method steps and apparatus for improving speech quality in
an audio signal received via microphone 302 using Linear Predictive Coding
(LPC) estimation 320, LPC analysis 326 and LPC synthesis 330.
As shown in Figure 3, an error signal is generated using LPC
estimation 320 from noise-suppressed audio signal 318. LPC coefficients
are generated using LPC analysis 326 of audio signal 306 received from
microphone 302, and is then applied to the error signal generated from LPC
estimation 320 to synthesize an output audio signal having improved speech
quality. Spec. ¶¶ 21–27.
Appeal 2020-004159
Application 14/636,562

3

Claims 1, 7, and 10 are independent. Representative claim 1 is
reproduced below with disputed limitations emphasized and bracketed
numerals added for clarity:
1. A method of improving speech quality in an audio
signal comprising speech and noise, after noise in the audio
signal is suppressed, the method comprising:
receiving from a microphone a first audio signal
comprising speech as captured by the microphone and noise as
captured by the microphone;
providing the first audio signal to a noise suppressor,
which is configured to suppress at least some of the noise in the
first audio signal and to thereby produce a noise-reduced version
of the first audio signal;
[1] using liner predictive coding (LPC) estimation to
generate an error signal from the noise-reduced version of the
first audio signal, the error signal comprising speech in the first
audio signal after at least some of the noise in the first audio
signal is removed;
generating linear predictive coding (LPC) coefficients by
performing linear predictive coding (LPC) analysis of the first
audio signal;
[2] applying the LPC coefficients generated by
performing LPC analysis of the first audio signal to the error
signal generated by LPC estimation of the noise reduced version
of the first audio signal to synthesize a second audio signal
having a reduced distortion speech and speech quality better
than the speech in the first audio signal; and
providing the second audio signal to an audio signal
transducer configured to produce audible sound waves from the
second audio signal.

Appeal Br. 27–32 (Claims App.).

Appeal 2020-004159
Application 14/636,562

4

REJECTION AND REFERENCES
Claims 1–13 stand rejected under 35 U.S.C. § 103 as obvious over the
combined teachings of Visser et al. (US 2013/0282373 A1; published Oct.
24, 2013; “Visser”) and Konchitsky (US 2014/0365212 A1; published Dec.
11, 2014). Final Act. 3–11.

ANALYSIS
In support of the obviousness rejection, the Examiner finds Visser
teaches most limitations of Appellant’s claim 1 and, similarly, claims 7 and
10, except for (1) “generating an error signal using linear predictive coding
(LPC) estimation . . . (2) generating [] LPC coefficients of the first audio
signal; and (3) applying the LPC coefficients to the error signal . . . to
synthesize a second audio signal.” Final Act. 3–4 (citing Visser ¶¶ 124, 132,
Figures 1, 7, and 8).
The Examine then relies on Konchitsky to teach the missing
limitations of Visser to support the conclusion of obviousness, i.e.,
it would have been obvious . . . to have a method of improving
speech quality by generating an error signal using the linear
predictive coding estimation method from an already noise-
reduced audio, synthesizing a second audio signal on the basis of
the first audio signal and error signal and providing the second
audio signal to a transducer to produce an improved sound signal
for the advantage of providing desirable techniques for
enchaining sound quality so that a receiver can receive a clear
and improve speech signal as taught by Konchitsky.

Id. at 5 (citing Konchitsky ¶¶ 13, 39, 40, 51, and Figure 8).

Appeal 2020-004159
Application 14/636,562

5

Appellant disputes the Examiner’s factual findings regarding both
Visser and Konchitsky. First, Appellant contends Konchitsky does not teach
or suggest the disputed limitations [1]–[2] of claim 1:
[1] “using liner predictive coding (LPC) estimation to
generate an error signal from the noise-reduced version of the
first audio signal, the error signal comprising speech in the first
audio signal after at least some of the noise in the first audio
signal is removed” [the “generating” limitation]; and

[2] “applying the LPC coefficients generated by performing
LPC analysis of the first audio signal to the error signal generated
by LPC estimation of the noise reduced version of the first audio
signal to synthesize a second audio signal having a reduced
distortion speech and speech quality better than the speech in the
first audio signal” [the “applying” limitation].

See Appeal Br. 8–10.
For example, Appellant argues:
Konchitsky only mentions an error signal in the background of
the invention section and in the context of explaining the
difference between the two methods of calculating LPC
coefficients. Konchitsky is otherwise silent with respect to
generating an error signal, let alone using LPC estimation to
generate an error signal from the noise-reduced version of the
first audio signal, as is explicitly recited in claim 1.
. . . .
Instead paragraph [0039] [of Konchitsky] discloses that the LPC
coefficients are calculated based on the components of first audio
signal buffer 302, which contains noise from the near- end
location. Paragraph [0040] discloses that instead of calculating
the LPC coefficients using Durbin-Levinson method, other
techniques such as covariance method, autocorrelation method,
or other methods may be used. And paragraph [0051] and Figure
8, item 814, of Konchitsky disclose calculating the LPC
coefficients of the first audio signal buffer, which contains noise
from the near-end location.
Appeal 2020-004159
Application 14/636,562

6

Appeal Br. 18–19 (citing Konchitsky ¶¶ 13, 39, 40, 51, and Figure 8)
(emphasis added).
Appellant also argues “Konchitsky does not disclose generating an
error signal and, consequently, does not disclose applying anything,
including LPC coefficients, to an error signal.” Id. at 20. According to
Appellant, Konchitsky only “discloses boosting selective frequencies of
audio signals from the far-end location when there is a lot of noise detected
by microphones at the near-end location, which, again, is very different from
what is explicitly recited in claim 1 [i.e., generating an error signal using
LPC estimation method from a noise-reduced audio signal . . . applying the
LPC coefficients to the error signal of the first audio signal to synthesize a
second audio signal].” Id. at 21.
Second, Appellant contends there would be no reason to combine
Visser and Konchitsky because:
Visser discloses voice-detector-based techniques for signal-level
matching between signals from multiple microphones, and
Konchitsky discloses boosting a signal received from a far-end
location based on detecting a noise level exceeding a threshold
at a near-end location.
Id. at 22–23.
The Examiner responds that Konchitsky teaches the disputed
limitations [1]–[2] of Appellant’s claim 1. Ans. 6. For example, the
Examiner takes the position that the cited paragraphs 12–13, 39, and Figure
3 of Konchitsky teach:
generating an error signal from applying a covariance method for
estimating a linear prediction coefficient for an audio signal . . .

Appeal 2020-004159
Application 14/636,562

7

applying linear prediction coefficients to the residual energy (or
error signal) when applying a covariance method for estimating
a linear prediction coefficient for an audio signal.

Ans. 6.
We are persuaded by Appellant’s arguments and disagree with the
Examiner’s position. Obviousness is a question of law based on underlying
factual findings (In re Baxter Int’l, Inc., 678 F.3d 1357, 1361 (Fed. Cir.
2012)), including what a reference teaches (In re Beattie, 974 F.2d 1309,
1311 (Fed. Cir. 1992)) and the existence of a reason to combine references
(In re Hyon, 679 F.3d 1363, 1365–66 (Fed. Cir. 2012)). At the outset, we
agree with Appellant that Konchitsky does not teach or suggest what the
Examiner finds. For example, as recognized by Appellant, Konchitsky’s
paragraphs 12–13 describe in the general background that (1) LPC is well-
known for speech analysis and (2) LPC coefficients can be estimated by two
widely used methods: autocorrelation and covariance. Appeal Br. 18; Reply
Br. 4 (citing Konchitsky ¶¶ 12–13). However, Konchitsky’s implementation
of the well-known LPC technique to improve the intelligibility in noisy
environments experienced in communication devices (e.g., mobile phones)
is different from the general concept of LPC discussed in the background
section (paragraphs 12–13) of Konchitsky as well as the specific
implementation of LPC recited in Appellant’s claimed invention.
For example, Konchitsky’s implementation of the well-known LPC
technique is shown in Figure 3 below:

Appeal 2020-004159
Application 14/636,562

8

Konchitsky’s Figure 3 shows a flow diagram of the processing of audio
signals, including: (1) a 1st audio signal (e.g., background noise) 302
received from microphones and (2) a 2nd audio signal (e.g., speech) 310
received from another communication device (e.g., mobile phone), where (3)
the LPC coefficients 304 and its LPC spectrum 306 are calculated based on
the 1st audio signal 302, and (4) the difference between the LPC spectrum
306 and the Fast Fourier Transform (FFT) 312 of the 2nd audio signal 310 is
calculated in order to boost the frequencies of the 2nd audio signal 310.
Konchitsky ¶¶ 38–41.
In contrast, Appellant’s claimed invention does not segregate audio
signals into background noise 310 and speech 310 as taught by Konchitsky.
Instead, Appellant’s claimed invention seeks to (1) suppress noise 312 from
an audio signal 306 received from microphone 302, as shown in Figure 3;
(2) use LPC estimation 320 to generate an error signal 322 from the noise-
reduced version of the audio signal; (3) generate LPC coefficients 328 by
Appeal 2020-004159
Application 14/636,562

9

performing LPC analysis 324 of the same audio signal 306, as shown in
Appellant’s Figure 3; and then (4) apply the LPC coefficients 328 to the
error signal 322 to synthesize an enhanced audio signal, as recited in
Appellant’s claim 1, and similarly, in claims 7 and 10.
Contrary to the Examiner’s findings, paragraphs 12–13, 39, and
Figure 3 of Konchitsky simply do not teach or suggest the disputed
limitations [1]–[2] of claim 1, including:
[1] “using liner predictive coding (LPC) estimation to
generate an error signal from the noise-reduced version of the
first audio signal, the error signal comprising speech in the first
audio signal after at least some of the noise in the first audio
signal is removed” [the “generating” limitation]; and

[2] “applying the LPC coefficients generated by performing
LPC analysis of the first audio signal to the error signal generated
by LPC estimation of the noise reduced version of the first audio
signal to synthesize a second audio signal having a reduced
distortion speech and speech quality better than the speech in the
first audio signal” [the “applying” limitation].

Appeal Br. 27. (Claims App.).
Because Konchitsky fails to teach the “using” and “applying”
limitations of Appellant’s claim 1, Konchitsky also fails to remedy the
deficiencies of Visser in order to arrive at Appellant’s claim 1.
For these reasons, we do not sustain the Examiner’s obviousness
rejection of independent claim 1 and, similarly, claims 7 and 10, and of their
respective dependent claims 2–6, 8, 9, and 11–13.

Appeal 2020-004159
Application 14/636,562

10

CONCLUSION
As such, we REVERSE the Examiner’s final rejection of claims 1–13
as obvious by Visser and Konchitsky.

DECISION SUMMARY
In summary:
Claim(s)
Rejected
35 U.S.C. § Reference(s)/
Basis
Affirmed Reversed
1–13 103 Visser,
Konchitsky
1–13

REVERSED