Ex Parte Reznik

Patent Trial and Appeal Board

Aug 15, 2016

13158013 (P.T.A.B. Aug. 15, 2016)

UNITED STA TES p A TENT AND TRADEMARK OFFICE
APPLICATION NO. FILING DATE
13/158,013 06/10/2011
15150 7590 08/17/2016
Shumaker & Sieffert, P, A,
1625 Radio Drive, Suite 100
Woodbury, MN 55125
FIRST NAMED INVENTOR
Yuriy Reznik
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.
1010-585US01/l 02107 8428
EXAMINER
CHBOUKI, TAREK
ART UNIT PAPER NUMBER
2165
NOTIFICATION DATE DELIVERY MODE
08/17/2016 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):
pairdocketing@ssiplaw.com
ocpat_uspto@qualcomm.com
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
BEFORE THE PATENT TRIAL AND APPEAL BOARD
Ex parte YURIY REZNIK
Appeal2014-009140
Application 13/158,013
Technology Center 2100
Before ERIC S. FRAHM, MELISSA A. RAAP ALA, and
JAMES W. DEJMEKAdministrative Patent Judges.
FRAHM, Administrative Patent Judge.
DECISION ON APPEAL
STATEMENT OF THE CASE
Introduction
Appellant appeals under 35 U.S.C. § 134(a) from a rejection of claims
1--48. We have jurisdiction under 35 U.S.C. § 6(b ).
We affirm-in-part.
Appeal2014-009140
Application 13/158,013
Appellant's Disclosed Invention
The invention relates "to image processing systems and, more
particularly, performing visual searches with image processing systems"
(Spec. if 2).
Exemplary Claims
Exemplary claims 1 and 3, reproduced below, with emphasis added, is
illustrative of the claimed subject matter:
1. A method for performing a visual search in a network system
in which a client device transmits query data via a network to a
visual search device, the method comprising:
extracting, with the client device, a set of image feature
descriptors from a query image, wherein the image feature
descriptors define at least one feature of the query image;
quantizing, with the client device, the set of image feature
descriptors at a first quantization level to generate first query data
representative of the set of image feature descriptors quantized
at the first quantization level;
transmitting, with the client device, the first query data to
the visual search device via the network;
determining, with the client device, second query data that
augments the first query data such that, when the first query data
is updated with the second query data, the updated first query
data is representative of the set of image feature descriptor
quantized at a second quantization level, wherein the second
quantization level achieves a more accurate representation of
the set of image feature descriptors than that achieved when the
set of image feature descriptors are quantized at the first
quantization level; and
transmitting, with the client device, the second query data
to the visual search device via the network to refine the first
query data.
3. The method of claim 1,
2
Appeal2014-009140
Application 13/158,013
wherein quantizing the image feature descriptors at a first
quantization level includes determining reconstruction points
such that the reconstruction points are each located at a center
of different ones of Voronoi cells defined for the image feature
descriptors, where the Voronoi cells include faces defining the
boundaries between the Voronoi cells and vertices where two or
more of the faces intersect,
wherein determining second query data includes:
determining additional reconstruction points such that the
additional reconstruction points are each located at a center of
each of the faces;
specifying the additional reconstruction points as offset
vectors from each of the previously determined reconstruction
points; and
generating the second query data to include the offset
vectors.
Examiner's Rejections
(1) The Examiner rejected claims 1, 2, 6-9, 14--17, 21-24, 29-32,
36-39 and 44--48 under§ 103(a) as being obvious over Gokturk et al. (US
2010/0166339 Al; published Jul. 1; 2010) and Nister et al. (US
2007/0214172 Al; published Sep. 13, 2007). Non-Final Act. 2-14.
(2) The Examiner rejected claims 3, 4, 10, 11, 18, 19, 25, 26, 33,
34, 40, and 41under§103(a) as being obvious over Gokturk, Nister, and
Sumakwell Muralla, A Method of Accelerating K-means by Directed
Perturbation of the Codevectors (July 1, 2006) (M.S. thesis, Oklahoma State
University)
(https:// shareok.org/bitstream/handle/11244/8207 /Muralla_okstate_0664 M_
1882.pdf) (hereinafter, "Muralla"), and Bessete et al. (US 2005/0285764 Al;
published Dec. 29, 2005). Non-Final Act. 14--20.
(3) The Examiner rejected claims 5, 12, 13, 20, 27, 28, 35, 42, and
43 under§ 103(a) as being obvious over Gokturk, Nister, and Murakami et
3
Appeal2014-009140
Application 13/158,013
al. (US 2005/0259884 Al; published Nov. 24, 2005). Non-Final Act. 21-
24.
Appellant's Contentions
( 1) Appellant contends (App. Br. 15-20) the Examiner erred in
rejecting claim 1 as being unpatentable under§ 103(a) as being obvious over
Gokturk and Nister because Gokturk in view ofNister fails to teach
"determining, with the client device, second query data that augments the
first query data such that, when the first query data is updated with the
second query data, the updated first query data is representative of the set of
image feature descriptor quantized at a second quantization level, wherein
the second quantization level achieves a more accurate representation of the
set of image feature descriptors than that achieved when the set of image
feature descriptors are quantized at the first quantization level;"
(2) Appellant contends (App. Br. 20-24) the Examiner erred in
rejecting claim 8 as being unpatentable under§ 103(a) as being obvious over
Gokturk and Nister because Gokturk in view ofNister fails to teach
"receiving, with the visual search device, second query data from the client
device via the network, wherein the second query data augments the first
data such that when the first query data is updated with the second query
data the updated first query data is representative of the set of image feature
descriptors quantized at a second quantization level, wherein the second
quantization level achieves a more accurate representation of the image
4
Appeal2014-009140
Application 13/158,013
feature descriptors than that achieved when quantizing at the first
quantization level;"
(3) Appellant contends (App. Br. 24--25) the Examiner erred in
rejecting claim 16 as being unpatentable under§ 103(a) as being obvious
over Gokturk and Nister for similar reasons as set forth with respect to claim
1.
(4) Appellant contends (App. Br. 25-26) the Examiner erred in
rejecting claim 23 as being unpatentable under§ 103(a) as being obvious
over Gokturk and Nister because Gokturk in view ofNister fails to teach
"receives second query data from the client device via the network, wherein
the second query data augments the first data such that when the first query
data is updated with the second query data the updated first query data is
representative of the image feature descriptors quantized at a second
quantization level";
(5) Appellant contends (App. Br. 26) the Examiner erred in rejecting
claim 31 as being unpatentable under§ 103(a) as being obvious over
Gokturk and Nister because Gokturk in view ofNister fails to teach "means
for determining second query data that augments the first query data such
that, when the first query data is updated with the second query data, the
updated first query data is representative of the set of image feature
descriptor quantized at a second quantization level";
(6) Appellant contends (App. Br. 27) the Examiner erred in rejecting
claim 38 as being unpatentable under§ 103(a) as being obvious over
Gokturk and Nister because Gokturk in view ofNister fails to teach "means
5
Appeal2014-009140
Application 13/158,013
for receiving second query data from the client device via the network,
wherein the second query data augments the first data such that when the
first query data is updated with the second query data the updated first query
data is representative of the set of image feature descriptors quantized at a
second quantization level,";
(7) Appellant contends (App. Br. 27-28) the Examiner erred in
rejecting claim 46 as being unpatentable under§ 103(a) as being obvious
over Gokturk and Nister because Gokturk in view ofNister fails to teach
"determine second query data that augments the first query data such that
when the first query data is updated with the second query data the updated
first query data is representative of the image feature descriptor quantized at
a second quantization level,";
(8) Appellant contends (App. Br. 28) the Examiner erred in rejecting
claim 47 as being unpatentable under§ 103(a) as being obvious over
Gokturk and Nister because Gokturk in view ofNister fails to teach
"wherein the second query data augments the first data such that when the
first query data is updated with the second query data the updated first query
data is representative of the image feature descriptor quantized at a second
quantization level,";
(9) Appellant contends (App. Br. 29) the Examiner erred in rejecting
claim 48 as being unpatentable under§ 103(a) as being obvious over
Gokturk and Nister because Gokturk in view ofNister fails to teach
"determines second query data that augments the first query data such that
when the first query data is updated with the second query data the updated
6
Appeal2014-009140
Application 13/158,013
first query data is representative of the image feature descriptor quantized at
a second quantization level,";
(10) Appellant contends (App. Br. 30) that the Examiner erred in
rejecting claims 2, 6, 9, 14, 17, 21, 24, 29, 32, 36, 39, and 44 for the same
reasons as claim 1 ;
(11) Appellant contends (App. Br. 30-32) the Examiner erred in
rejecting claims 7, 15, 22, 30, 37, and 45 as being unpatentable under
§ 103(a) as being obvious over Gokturk and Nister because Gokturk in view
ofNister fails to teach "determining third query data that further augments
the first and second query data such that when the first query data after being
augmented by the second query data is updated with the third query data the
successively updated first query data is representative of the image feature
descriptors quantized at a third quantization level, wherein the third
quantization level achieves an even more accurate representation of the
image feature descriptor data than that achieved when quantizing at the
second quantization level, and transmitting the third query data to the visual
search device via the network to successively refine the first query data after
being augmented by the second query data" as recited by claim 7; 1
(12) Appellant contends (App. Br. 32-34) the Examiner erred in
rejecting claims 3, 4, 10, 11, 18, 19, 25, 26, 33, 34, 40, and 41 as being
unpatentable under§ 103(a) as being obvious over Gokturk, Nister, Muralla,
1 Based on Appellant's argument (App. Br. 30-32) with regard to the
§ 103(a) rejection of claims 7, 15, 22, 30, 37, and 45 as being obvious over
Gokturk and Nister we select claim 7 as representative of the group of
claims.
7
Appeal2014-009140
Application 13/158,013
and Bessette because the combination of Gokturk, Nister, Muralla, and
Bessette fails to teach "quantizing the image feature descriptors at a first
quantization level includes determining reconstruction points such that the
reconstruction points are each located at a center of different ones of
V oronoi cells defined for the image feature descriptors, where the V oronoi
cells include faces defining the boundaries between the V oronoi cells and
vertices where two or more of the faces intersect" and "determining second
query data includes: determining additional reconstruction points such that
the additional reconstruction points are each located at a center of each of
the faces, specifying the additional reconstruction points as offset vectors
from each of the previously determined reconstruction points, and
generating the second query data to include the offset vectors"; 2
(13) Appellant contends (App. Br. 35) that the Examiner erred in
rejecting claims 5, 12, 13, 20, 27, 28, 35, 42, and 43 for the same reasons as
claim 1. 3
Issues on Appeal
(1) Did the Examiner err in rejecting claims 1, 2, 6, 8, 9, 14, 16, 17,
21, 23, 24, 29, 31, 32, 36, 38, 39, 44, and 46-48 as being unpatentable under
35 U.S.C. § 103(a) as obvious in light of Gokturk in view ofNister as
2 Based on Appellant's argument (App. Br. 35) with regard to the§ 103(a)
rejection of claims 3, 4, 10, 11, 18, 19, 25, 26, 33, 34, 40, and 41 as being
obvious over Gokturk, Nister, Muralla, and Bessette, we select claim 3 as
representative of the group of claims.
3 Based on Appellant's argument (App. Br. 35) with regard to the§ 103(a)
rejection of claims 5, 12, 13, 20, 27, 28, 35, 42, and 43 as being obvious
over Gokturk, Nister, and Murakami, we select claim 5 as representative of
the group of claims.
8
Appeal2014-009140
Application 13/158,013
teaching "determines second query data that augments the first query data
such that when the first query data is updated with the second query data the
updated first query data is representative of the image feature descriptors
quantized at a second quantization level," as required by representative
claim 1 ?4
(2) Did the Examiner err in rejecting claims 7, 15, 22, 30, 37, and 45
as being unpatentable under 35 U.S.C. § 103(a) as obvious in light of
Gokturk in view ofNister as teaching "determining third query data that
further augments the first and second query data such that when the first
query data after being augmented by the second query data is updated with
the third query data the successively updated first query data is
representative of the image feature descriptors quantized at a third
quantization level, wherein the third quantization level achieves an even
more accurate representation of the image feature descriptor data than that
achieved when quantizing at the second quantization level, and transmitting
the third query data to the visual search device via the network to
successively refine the first query data after being augmented by the second
query data," as required by representative claim 7?
(3) Did the Examiner err in rejecting claims 3, 4, 10, 11, 18, 19, 25,
26, 33, 34, 40, and 41 as being unpatentable under 35 U.S.C. § 103(a) as
obvious in light of Gokturk, Nister, Muralla, and Bessette as teaching
4 Appellant presents duplicative arguments in contentions 1-10 for slight
variations of a limitation present in each independent claims 1, 8, 16, 23, 31,
38, and 46-48 of which we select claim 1 as representative. Further, the
obviousness rejection of claims 5, 12, 13, 20, 27, 28, 35, 42, and 43 under
§ 103(a) over Gokturk, Nister, and Murakami will respectively stand or fall
for the same reasons as representative claim 1 as discussed with the
obviousness rejection of claims 1, 8, 16, 23, 31, 38, and 46-48.
9
Appeal2014-009140
Application 13/158,013
"quantizing the image feature descriptors at a first quantization level
includes determining reconstruction points such that the reconstruction
points are each located at a center of different ones of Voronoi cells defined
for the image feature descriptors, where the Voronoi cells include faces
defining the boundaries between the V oronoi cells and vertices where two or
more of the faces intersect" and "determining second query data includes:
determining additional reconstruction points such that the additional
reconstruction points are each located at a center of each of the faces,
specifying the additional reconstruction points as offset vectors from each of
the previously determined reconstruction points, and generating the second
query data to include the offset vectors," as required by representative claim
3?
ANALYSIS
We have reviewed the Examiner's rejections (Non-Final Act. 2-24) in
light of Appellant's contentions in the Appeal Brief (App. Br. 13-35) and
the Reply Brief (Reply Br. 3-18) that the Examiner has erred, as well as the
Examiner's response to Appellant's arguments in the Appeal Brief (see Ans.
2-5). We disagree with Appellant's first through eleventh contentions listed
above as to the obviousness rejection over Gokturk and Nister. We agree
with Appellant's twelth contention listed above as to the obviousness
rejection over Gokturk, Nister, Muralla, and Bessete.
10
Appeal2014-009140
Application 13/158,013
Claims 1, 2, 6, 8, 9, 14, 16, 17, 21, 23, 24, 29, 31, 32,
36, 38, 39, 44, and 46--48
We disagree with Appellant's conclusions regarding the obviousness
rejection of claims 1, 2, 6, 8, 9, 14, 16, 17, 21, 23, 24, 29, 31, 32, 36, 38, 39,
44, and 46-48 over Gokturk and Nister. We adopt as our own ( 1) the
findings and reasons set forth by the Examiner in the action from which this
appeal is taken (Non-Final Act. 3--4), and (2) the reasons set forth by the
Examiner in the Examiner's Answer in response to Appellant's Appeal Brief
(Ans. 2-3). We highlight and amplify certain teachings and suggestions of
Gokturk and Nister as follows.
Gokturk discloses a search operation "to identify one or more items
that have a visual characteristic that satisfies at least some of the one or more
search criteria" (see Abs.) with a multi-pass cascaded search where each
subsequent pass uses greater detail (see Gokturk, i-fi-f 181-182). The
Examiner finds that by describing a first pass of global features and a second
pass of local features Gokturk teaches "second query data that augments the
first query data" as required by representative claim 1 (see Ans. 3, Gokturk
i1255). We agree with the Examiner that Gokturk and Nister teaches or
suggests the subject matter of representative claim 1, including teaching
"determines second query data that augments the first query data such that
when the first query data is updated with the second query data the updated
first query data is representative of the image feature descriptors quantized at
a second quantization level." (Ans. 2-3).
In view of the foregoing, we sustain the rejection based on Gokturk
and Nister of representative claim 1, as well as claims 2, 6, 8, 9, 14, 16, 17,
11
Appeal2014-009140
Application 13/158,013
21, 23, 24, 29, 31, 32, 36, 38, 39, 44, and 46-48 grouped therewith. (See
supra note 4).
Claims 7, 15, 22, 30, 37, and 45
We disagree with Appellant's conclusions regarding the obviousness
rejection of claims 7, 15, 22, 30, 37, and 45 over Gokturk and Nister. We
adopt as our own ( 1) the findings and reasons set forth by the Examiner in
the action from which this appeal is taken (Non-Final Act. 6), and (2) the
reasons set forth by the Examiner in the Examiner's Answer in response to
Appellant's Appeal Brief (Ans. 3--4). We highlight and amplify certain
teachings and suggestions of Gokturk and Nister as follows.
As discussed supra, Gokturk discloses a visual search operation with
a multi-pass cascaded search where each subsequent pass uses greater detail.
Gokturk's disclosure makes it clear that more than two passes may be
beneficial to increase the quality of visual search. (see Gokturk, i-fi-f 181-
182). Accordingly, we agree with the Examiner that Gokturk and Nister
teach or suggest the subject matter of representative claim 7, including
teaching "determining third query data that further augments the first and
second query data such that when the first query data after being augmented
by the second query data is updated with the third query data the
successively updated first query data is representative of the image feature
descriptors quantized at a third quantization level, wherein the third
quantization level achieves an even more accurate representation of the
image feature descriptor data than that achieved when quantizing at the
second quantization level; and transmitting the third query data to the visual
search device via the network to successively refine the first query data after
being augmented by the second query data." (Ans. 3--4).
12
Appeal2014-009140
Application 13/158,013
In view of the foregoing, we sustain the rejection based on Gokturk
and Nister of representative claim 7, as well as claims 7, 15, 22, 30, 37, and
45 grouped therewith.
Claims 3, 4, 10, 11, 18, 19, 25, 26, 33, 34, 40, and 41
The Examiner finds the combination of Gokturk, Nister, Muralla, and
Bessette teaches all the limitations of claim 3, including that Muralla teaches
"wherein quantizing the image feature descriptors at a first quantization
level includes determining reconstruction points such that the reconstruction
points are each located at a center of different ones of Voronoi cells defined
for the image feature descriptors, where the Voronoi cells include faces
defining the boundaries between the V oronoi cells and vertices where two or
more of the faces intersect, wherein determining second query data includes:
determining additional reconstruction points such that the additional
reconstruction points are each located at a center of each of the faces;" and
"specifying the additional reconstruction points as offset vectors from each
of the previously determined reconstruction points; and generating the
second query data to include the offset vectors." (Non-Final Act. 14--16).
Appellant contends "Gokturk in view ofNister, [Muralla] and Bassette [sic]
does not teach or suggest determining second query data includes
determining additional reconstruction points such that the additional
reconstruction points are each located at a center of each of the faces, as
recited by claim 3." (App. Br. 33). Appellant states the use of Voronoi cells
themselves is not novel, "but contends that the use of Voronoi cells to
identify second query data that augments first query data so as to achieve a
more accurate representation of the image feature descriptors represented by
the first query data, as recited in the claims, is novel and non-obvious in the
13
Appeal2014-009140
Application 13/158,013
context of a cascading visual search." (Reply Br. 15). We agree with
Appellant's conclusions regarding the obviousness rejection of claims 3, 4,
10, 11, 18, 19, 25, 26, 33, 34, 40, and 41 over Gokturk, Nister, Muralla, and
Bessette.
Muralla discloses a method for data clustering as related to image
analysis for data mining (Muralla 28) and "[Stochastic Relaxation Scheme]
is an attempt to reduce the combinatorial complexity of the [Simulated
Annealing] but still provide a close to globally optimal solution." (Muralla
15). Further,
In the proposed method, perturbations are added to the
codevectors not in any random direction and amount but in a
more directed way towards the general direction where the
codevectors are going. Just like in case of the standard K-Means
algorithm, the convergence using the new method also proceeds
in a monotonically decreasing manner.
(Muralla 1 7).
However; the Examiner has not shown Muralla uses "wherein quantizing the
image feature descriptors at a first quantization level includes determining
reconstruction points such that the reconstruction points are each located at a
center of different ones of V oronoi cells defined for the image feature
descriptors, where the V oronoi cells include faces defining the boundaries
between the Voronoi cells and vertices where two or more of the faces
intersect, wherein determining second query data includes: determining
additional reconstruction points such that the additional reconstruction
points are each located at a center of each of the faces." Rather, Muralla
provides a general framework for the use of V oronoi cells through an
algorithm fork-means clustering (Muralla 9-10), but there is no nexus tying
Muralla's k-means clustering to the claimed cascaded visual search. While
14
Appeal2014-009140
Application 13/158,013
Muralla teaches the general concept of an improved k-means clustering, the
Examiner fails to show how k-means clustering as taught by Muralla would
be applied to the cascaded visual search as taught by Gokturk and Nister to
read on the limitation of claim 3, which requires the use of reconstruction
points at a center of different Voronoi cells defined for the image feature
descriptors in a cascaded visual search. The Examiner's rejection cites
solely to Muralla for this limitation (Non-Final 14--15).
The Examiner responds that Muralla also "teaches the cluster centers
being codebook/codevectors and determining to be located around the data
points" (Ans. 4--5; Muralla 4, 15, and 23). However, the Examiner does not
explain how this k-means clustering algorithm relates to the use of
reconstruction points at a center of different V oronoi cells defined for the
image feature descriptors in a cascaded visual search discussed above (See
Ans. 5). Accordingly, we are persuaded the Examiner has failed to show
Muralla teaches or suggests the disputed limitations of claim 3.
We are, therefore, constrained by the record to find the Examiner
erred in rejecting claim 3, as well as claims 4, 10, 11, 18, 19, 25, 26, 33, 34,
40, and 41 grouped therewith.
Claims 5, 12, 13, 20, 27, 28, 35, 42, and 43
Appellant has failed to show that the Examiner erred in determining
the combination of Gokturk, Nister, and Murakami teaches or suggests the
additional limitations as recited in claims 5, 12, 13, 20, 27, 28, 35, 42, and
43 because Appellant does not address the merits of the inclusion of
Murakami in the combination, or otherwise present arguments on the merits
with regard to these claims (see App. Br. 35). See 37 C.F.R.
§ 41.37(c)(l)(iv) (requiring a statement in the briefs as to each ground of
15
Appeal2014-009140
Application 13/158,013
rejection presented by Appellant for review and that arguments not
presented in the briefs by Appellant will be refused consideration). As such,
Appellant has not separately argued the Examiner erred in rejecting claims
5, 12, 13, 20, 27, 28, 35, 42, and 43 or otherwise shown this obviousness
rejection to be in error. See id.
Accordingly, we sustain the obviousness rejection of claims 5, 12, 13,
20, 27, 28, 35, 42, and 43 (based on Gokturk, Nister, and Murakami) for the
same reasons as claim 1 for Gokturk and Nister discussed supra.
CONCLUSIONS
(1) The Examiner did not err in rejecting claims 1, 2, 6-9, 14--17, 21-
24, 29-32, 36-39 and 44--48 under§ 103(a) as being unpatentable by
Gokturk and Nister because the combination of Gokturk and Nister teaches
or suggests each of the disputed limitations in representative claim 1.
(2) The Examiner erred in rejecting claims 3, 4, 10, 11, 18, 19, 25, 26,
33, 34, 40, and 41 under§ 103(a) as being unpatentable by Gokturk, Nister,
Muralla, and Bessette because the Examiner has failed to show Muralla
teaches or suggests a k-means clustering algorithm related to the use of
reconstruction points at a center of different V oronoi cells defined for the
image feature descriptors in a cascaded visual search, as required in
representative claim 3.
(3) Appellant has not shown the Examiner erred in rejecting claims 5,
12, 13, 20, 27, 28, 35, 42, and 43 under 35 U.S.C. § 103(a) as being
unpatentable over the combination of Gokturk, Nister, and Murakami.
16
Appeal2014-009140
Application 13/158,013
DECISION
We affirm the Examiner's rejections of claims 1, 2, 5-9, 12-17, 20-
24, 27-32, 35-39 and 42--48, and reverse the Examiner's rejections of
claims 3, 4, 10, 11, 18, 19, 25, 26, 33, 34, 40, and 41.
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
17