Ex Parte Davis

Patent Trial and Appeal Board

Aug 1, 2016

13000515 (P.T.A.B. Aug. 1, 2016)

UNITED STA TES p A TENT AND TRADEMARK OFFICE
APPLICATION NO. FILING DATE
13/000,515 12/21/2010
23117 7590 08/03/2016
NIXON & V ANDERHYE, PC
901 NORTH GLEBE ROAD, 11 TH FLOOR
ARLINGTON, VA 22203
FIRST NAMED INVENTOR
Andrew G. Davis
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.
RYM-36-2357 4072
EXAMINER
HUANG, JEN-SHI
ART UNIT PAPER NUMBER
2423
NOTIFICATION DATE DELIVERY MODE
08/03/2016 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):
PTOMAIL@nixonvan.com
pair_nixon@firsttofile.com
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
BEFORE THE PATENT TRIAL AND APPEAL BOARD
Ex parte ANDREW G. DA VIS
Appeal2015-003936
Application 13/000,515
Technology Center 2400
Before JOHN A. JEFFERY, BRADLEY W. BAUMEISTER, and
DENISE M. POTHIER, Administrative Patent Judges.
JEFFERY, Administrative Patent Judge.
DECISION ON APPEAL
Appellant appeals under 35 U.S.C. § 134(a) from the Examiner's
decision to reject claims 1---6, 10-15, and 17-20. The remaining, pending
claims are objected to as being dependent upon a rejected based claim. We
have jurisdiction under 35 U.S.C. § 6(b). We reverse.
Appeal2015-003936
Application 13/000,515
STATEMENT OF THE CASE
Appellant's invention measures a compressed video's quality.
Spec. 1. Encoding and decoding a video signal degrades its quality. Id.
Measuring this degradation by referencing the original signal is often
inconvenient. Id. Accordingly, Appellant's method is a no-reference,
decoder-based assessment. Id. at 3. That is, the method does not reference
the original signal, and the method operates inside a decoder to estimate the
decoded video's quality. Id. As part of this estimate, the invention
generates a measurement that is a function of the signal's prediction
residual. Id. at 16. The prediction residual, in predictive coding, is the
difference between the actual signal and the corresponding predicted signal.
Id. at 1.
Claim 1, reproduced below with our emphasis, is illustrative:
1. A method of generating a measure of quality for a video
signal representative of a plurality of frames, the video signal
having: an original form; an encoded form in which the video
signal has been encoded using a compression algorithm utilizing
a variable quantizer step size such that the encoded signal has a
quantizer step size parameter associated therewith and utilizing
differential coding such that the encoded signal contains
representations of the prediction residual of the signal; and a
decoded form in which the encoded video signal has been at least
in part reconverted to the original form, the method comprising:
a) generating a first quality measure which is dependent
on said quantizer step size parameter according to a
predetermined relationship;
b) generating a masking measure, the masking measure
being dependent on the spatial complexity of at least part of the
frames represented by the video signal in the decoded form
according to a predetermined relationship; and
2
Appeal2015-003936
Application 13/000,515
c) generating a combined measure, the combined measure
being dependent upon both the first measure and the masking
measure according to a predetermined relationship;
wherein the method also includes
generating a second measure which is dependent on the
prediction residual of the signal according to a predetermined
relationship;
identifying one or more regions of the picture for which
the second measure exceeds a threshold;
and wherein the masking measure is dependent on the
spatial complexity of the identified one or more reg10ns
according to a predetermined relationship.
THE REJECTION
The Examiner rejected claims 1---6, 10-15, and 17-20 under 35 U.S.C.
§ 103(a) as unpatentable over Davis1 (US 2008/0317111 Al; published
1 \Ve note that Davis, in fact, does not qualify as prior art. The reference and
the application in this appeal have the same inventive entity-Andrew G.
Davis. Accordingly, the Davis reference does not qualify as prior art under
35 U.S.C. § 102(a) or (e) because it is not "by others." See MPEP §§ 2132
(III), 2136 (9th ed. Rev. 07.2015, Nov. 2015). Furthermore, Appellant's
application is a U.S. National Stage Application. See Notice of Acceptance
of Application under 35 U.S.C. 371 and 37 C.F.R. 1.495, dated January 25,
2011. So, Appellant's effective filing date is the April 29, 2009,
international filing date. See 35 U.S.C. § 363; see also MPEP § 1893.03(b).
Appellant's international filing date of April 23, 2009, should not be
confused with the date of entry into the national stage-December 21, 2010.
See MPEP § 1893.03(b). So, Davis also does not qualify as prior art under
35 U.S.C. § 102(b) because the reference was published on December 25,
2008-less than a year before Appellant's effective filing date of April 23,
2009.
Nevertheless, Davis is also a National Stage Application with a
corresponding international publication dated June 14, 2007-more than a
year before Appellant's filing date. This corresponding international
3
Appeal2015-003936
Application 13/000,515
Dec. 25, 2008) and Bhaskaran (US 2004/0151243 Al; published
Aug. 5, 2004). Final Act. 4--9.2
CONTENTIONS
The Examiner finds that Davis discloses every recited element of
claim 1 except for the recited identifying regions where a measure exceeds
the threshold, but relies on Bhaskaran as teaching this feature in concluding
the claim would have been obvious. Final Act. 5. In particular, the
Examiner finds that Bhaskaran uses a threshold to determine high and low
activity blocks. Id. (citing Bhaskaran i-f 32, Fig. 2); Ans. 6 (citing
B haskaran i-fi-1 3 5, 41).
Regarding the recited generating a second measure, the Examiner
finds that Davis uses prediction residuals as a quality measure. Ans. 3---6.
According to the Examiner, Davis discloses prediction residuals in
paragraphs 4, 5 and 26. Id. at 3-5. Furthermore, the Examiner finds that the
MPEG3 standard uses prediction residuals. Adv. Act.; Ans. 3 (citing
Davis i-f 16).
publication is materially equivalent to the Davis publication relied upon by
the Examiner. Accordingly, we understand the rejection to be based on
Davis's international publication-not the national stage publication. For
clarity, though, the citations in this opinion refer to the U.S. Patent
Application Publication so as to be consistent with the Examiner's and
Appellant's citations.
2 Throughout this opinion, we refer to (1) the Final Action mailed January
28, 2014 ("Final Act."); (2) the Advisory Action mailed April 24, 2014
("Adv. Act."); (3) the Appeal Brief filed August 20, 2014 ("App. Br.");
(4) the Examiner's Answer mailed December 17, 2014 ("Ans."); and (5) the
Reply Brief filed February 18, 2015 ("Reply Br.").
3 MPEG stands for Moving Picture Expert Group. Bhaskaran i-f 30.
4
Appeal2015-003936
Application 13/000,515
Among other arguments, Appellant argues that Davis does not
generate a second measure that is dependent on a prediction residual, as
recited. App. Br. 9-11; Reply Br. 6-8. In particular, Appellant contends
that Davis only predicts the quality that a human would attribute to the
image. Reply Br. 6. According to Appellant, Davis's video-quality
prediction is different from a prediction residual. Id. In Appellant's view, a
frame's prediction residual is understood in the art to mean the difference
between an extrapolated frame and the required frame. Id. at 7 (citing
Spec. 1-2). Appellant further argues that the prediction residual indicates
how noticeable the errors are when the view is changing rapidly.
Reply Br. 8.
Furthermore, Appellant contends that Davis only uses prediction
residuals to generate one frame from another, not for measuring quality.
App. Br. 10 (citing Adv. Act.); Reply Br. 5.
ISSUE
Under§ 103, has the Examiner erred in rejecting claim 1 by finding
that Davis and Bhaskaran would have taught or suggested the recited second
measure dependent on the prediction residual?
ANALYSIS
We begin by construing the key disputed limitation of claim 1, which
recites, in pertinent part, a "prediction residual." To this end, we look to the
Specification, which "acts as a dictionary when it expressly defines terms
used in the claims or when it defines terms by implication." Phillips v. AWH
Corp., 415 F.3d 1303, 1321 (Fed. Cir. 2005) (en bane) (citations omitted).
5
Appeal2015-003936
Application 13/000,515
The Specification states that the difference between a signal and the
corresponding predicted signal is known as the prediction residual. Spec. 1-
2. Based on this discussion (id.), Appellant has, in effect, defined implicitly
the term "prediction residual" to at least require a difference involving a
corresponding predicted signal.
With this construction, we tum to the cited prior art. In the rejection,
the Examiner cites Bhaskaran's Figure 2 and paragraph 32 for teaching the
recited second measure dependent on the prediction residual. Final Act. 5.
But in the Answer, the Examiner cites Davis's paragraphs 4, 5, and 26 as
corresponding to the recited prediction residual. Ans. 4---6.
Appellant argues that the Examiner shifts positions on whether Davis
teaches the prediction residual. App. Br. 7-8 (citing Final Act. 5); Reply Br.
2-3. To the extent there is ambiguity on this point, the Examiner, has
nevertheless made clear on this record that Davis is relied on for disclosing
the recited prediction residual and second measure. See Adv. Act; Ans. 4--6.
So, consistent with the Examiner's intent, we confine our discussion to
Davis.
In the cited paragraphs, Davis discusses using a quantizer's step size
to measure a video signal's quality. Davis i-fi-1 4--5, cited in Ans. 4---6.
Specifically, Davis's compression algorithm uses a variable-quantizer step
size to encode a video signal. Davis i14. Davis explains that many encoding
schemes transmit the step size as a parameter. Id. i1 5. Instead of
recalculating the step size each time to predict video quality, Davis can use
the transmitted parameter. Id. i15 (emphasis added). Davis further explains
that picture complexity can also be taken into account to predict a viewer's
subjective degree of quality. Id. i126. That is, Davis uses both quantizer
6
Appeal2015-003936
Application 13/000,515
step size and complexity measurements to measure video quality. See
id. ff 5, 26; see also id. Abstract.
We agree with Appellant's argument (Reply Br. 6) that Davis's
quality prediction (Davis i-f 5) differs from the recited prediction residual.
Davis's video-quality prediction (id. i-f 5) refers to a prediction about
viewer-perceived distortion. See id. i-f 3 (discussing the purpose of the
quality metric). That is, the Examiner has only identified a prediction about
video quality, not a predicted signal-let alone a difference with a predicted
signal. See Ans. 4---6.
The Examiner's explanation lacks a discussion of differences with
predicted frames. See Final Act. 4--5; Ans. 4---6. To be sure, Davis discloses
pixel differences. See, e.g., Davis i-fi-122-24. But instead of using the
difference between the actual frame and the predicted frame, Davis uses a
difference of pixels in a single image to obtain a contrast measure. See
id. i-f 22; see also Davis, Fig. 2 (showing the pixels involved in calculating
Davis's pixel differences); accord Reply Br. 7. Accordingly, we agree with
Appellant (Reply Br. 6-7) that the Examiner-cited predictions in Davis
(Ans. 4---6) are not prediction residuals, as required by claim 1.
Furthermore, to the extent that the Examiner finds that Davis's step
size corresponds to the prediction residual (see Ans. 4), we find at least two
problems with such a correspondence.
First, the step size is a measure of the quantization granularity.
Davis i-f 2. For example, encoding techniques quantize the magnitude of a
transform-coefficient series. Id. The step size represents the level of this
quantization. Id. But the Examiner does not explain how this quantization
granularity involves a predicted signal or a difference. See Ans. 4.
7
Appeal2015-003936
Application 13/000,515
Second, the claim also recites a step size in step (a). And the
Examiner finds that step (a) corresponds to Davis's unit 3. Final Act. 4.
Unit 3 averages the quantizer step-size for a picture. Davis i-f 19. But the
Examiner does not explain (1) how Davis's step size can be both step size
and prediction residual, or (2) how Davis's step size can be used in both the
recited first and second measures. See Ans. 4. Given this unresolved
inconsistency, the Examiner has not shown that Davis's step size
corresponds to the recited prediction residual. See id.
The Examiner also explains that the MPEG standard uses prediction
residuals. Adv. Act.; Ans. 3 (citing Davis i-f 16). Likewise, Appellant
acknowledges that Davis uses prediction residuals to generate frames. See
App. Br. 10. The Examiner, however, does not identify a second measure
dependent on these residuals. See Ans. 3. Nor does the Examiner explain
whether this dependency is in accordance with the recited predetermined
relationship. See id. Rather, Davis explicitly states that one version of the
bit-stream analysis does not require MPEG's predicted-macroblock
reconstruction. Davis i-f 44, quoted in App. Br. 9-10. Accordingly, we are
persuaded by Appellant (App. Br. 9-11; Reply Br. 6-8) that Davis lacks
using the prediction residual, as recited in claim 1.
Therefore, we do not sustain the rejections of (1) independent claim 1;
(2) independent claim 14, which also requires a prediction residual; or
(3) dependent claims 2---6, 10-13, 15, and 17-20 for similar reasons.
Because this issue is dispositive regarding the Examiner's error in rejecting
these claims, we need not address Appellant's other arguments.
8
Appeal2015-003936
Application 13/000,515
CONCLUSION
The Examiner erred in rejecting claims 1---6, 10-15, and 17-20 under
§ 103.
DECISION
The Examiner's decision rejecting claims 1---6, 10-15, and 17-20 is
reversed.
REVERSED
9