Ex Parte KE et al

Patent Trial and Appeal Board

Jul 29, 2016

12610810 (P.T.A.B. Jul. 29, 2016)

UNITED STA TES p A TENT AND TRADEMARK OFFICE
APPLICATION NO. FILING DATE FIRST NAMED INVENTOR
12/610,810 11102/2009
45809 7590 08/02/2016
SHOOK, HARDY & BACON L.L.P.
(MICROSOFT TECHNOLOGY LICENSING, LLC)
INTELLECTUAL PROPERTY DEPARTMENT
2555 GRAND BOULEVARD
KANSAS CITY, MO 64108-2613
QIFAKE
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.
328346.01/MFCP.152519 1213
EXAMINER
CHOI, YUK TING
ART UNIT PAPER NUMBER
2164
NOTIFICATION DATE DELIVERY MODE
08/02/2016 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):
IPDOCKET@SHB.COM
IPRCDKT@SHB.COM
usdocket@microsoft.com
PTOL-90A (Rev. 04/07)
UNITED STATES PATENT AND TRADEMARK OFFICE
BEFORE THE PATENT TRIAL AND APPEAL BOARD
Ex parte QIF A KE, MING LIU, and YI LI
Appeal2015-002174
Application 12/610,810
Technology Center 2100
Before MAHSHID D. SAADAT, JOHNNY A. KUMAR, and
JON M. JURGOV AN, Administrative Patent Judges.
JURGOVAN, Administrative Patent Judge.
DECISION ON APPEAL
Appellants 1 seek review under 35 U.S.C. § 134(a) from a Final
Rejection of claims 1-16 and 18-20. Claim 17 was canceled. We have
jurisdiction under 35 U.S.C. § 6(b).
We affirm.2
1 Appellants identify Microsoft Corporation as the real party in interest. (Br.
3.)
2 Our Decision refers to the Specification (filed Nov. 2, 2009) ("Spec."), the
Final Office Action (mailed Mar. 7, 2014) ("Final Act."), the Appeal Brief
(filed Sept. 8, 2014) ("Br."), and the Examiner's Answer (mailed Oct. 3,
2014) ("Ans.").
Appeal2015-002174
Application 12/610,810
STATEMENT OF THE CASE
The claims are directed to content-based image searching. (Spec. i-f 5.)
Claim 1, reproduced below, is illustrative of the claimed subject matter:
1. A computer-implemented method for searching a plurality
of images, the method comprising:
receiving a search query that includes an image;
identifying, by a computing device, a plurality of first
descriptor identifiers based on the search query, each of the first
descriptor identifiers comprising an identifier used to identify a
respective descriptor, each descriptor having been calculated
over a respective portion of the image comprising a subset of
pixels of the image that includes a respective interest point in the
image;
searching a plurality of indexed images by comparing one
or more of the first descriptor identifiers to one or more second
descriptor identifiers associated with each of the indexed images;
identifying one or more indexed images that each has a
threshold number of descriptor identifiers that match one or more
of the first descriptor identifiers; and
ranking the one or more indexed images based on the
companson.
REJECTIONS
Claims 1-12, 14--16, 18, and 19 stand rejected under 35 U.S.C.
§ 103(a) based on Li (US 2007/0067345 Al; publ. Mar. 22, 2007), Shiiyama
(US 2003/0026476 Al; publ. Feb. 6, 2003), and Erol (US 2008/0027983 Al;
publ. Jan. 31, 2008). (Final Act. 3-17.)
Claim 13 stands rejected under 35 U.S.C. § 103(a) based on Li,
Shiiyama, Erol, and Gururajan (US 2007 /0077987 Al; publ. Apr. 5, 2007).
(Final Act. 17 .)
2
Appeal2015-002174
Application 12/610,810
Claim 20 stands rejected under 35 U.S.C. § 103(a) based on Li,
Shiiyama, Erol, and Valz (US 2008/0154798 Al; publ. June 26, 2008).
(Final Act. 18-22.)
ANALYSIS
IdentifYing a Plurality of Descriptor Identifiers
Regarding claim 1, Appellants contend the combination of Li,
Shiiyama, and Erol does not teach identifying a set of descriptor identifiers
for an image, in which each descriptor identifier identifies a corresponding
descriptor that was calculated over a respective portion of the image
comprising a subset of pixels of the image. (Br. 6-9.) Appellants argue Li's
feature vector and Shiiyama's color descriptor values include a collection of
numerical features representing an image, but each of the numerical features
in the vector or color descriptor does not correspond to a respective portion
of the image, thus Li and Shiiyama do not teach the claimed plurality of
descriptor identifiers. (Br. 7-8.) Appellants contend Erol fails to cure the
alleged deficiencies of Li and Shiiyama, because the descriptors of Erol
require a calculation to identify matches whereas the claimed descriptor
identifiers are integer numbers that provide simple integer lookup for easier
image matching during a search process. (Br. 8-9 and 11-12.) Appellants
argue claims 10 and 20 on a basis similar to these arguments for claim 1.
(Br. 14--16 and 17-18.)
Appellants' argument that the claimed descriptor identifiers are simple
integers is not commensurate with the scope of claim 1, which does not
require any specific form for the descriptor identifiers. Under the broadest
reasonable interpretation of "descriptor identifier" that is consistent with
3
Appeal2015-002174
Application 12/610,810
Appellants' disclosure, we agree with the Examiner's finding that the labels
given to identify Li's descriptors (Ans. 5 (citing Li i-fi-122, 24, 33; see also
i1 34: feature vector descriptors identified as RGB color histogram and HSV
color histogram)) and Shiiyama's descriptors (Ans. 6-7 (citing Shiiyama ,-r
125: color descriptors identified as YCoeff, CbCoeff, and CrCoeff)) teaches
the claimed descriptor identifiers that are compared to search for matching
images. (See In re Morris, 127 F.3d 1048, 1054 (Fed. Cir. 1997)). Further,
because the proposed rejection is based on the combination of the
references, Appellants' challenge to the references individually is not
persuasive of error in the Examiner's rejection. (See In re Keller, 642 F .2d
413, 426 (CCP A 1981) ("[O]ne cannot show non-obviousness by attacking
references individually where ... the rejections are based on combinations
of references").) The Examiner finds, and we agree, that the method of
searching images using the descriptor identifiers of Li and Shiiyama (Ans.
5-7), combined with the method of calculating a descriptor over a respective
portion of an image comprising a subset of pixels of the image of Erol (Ans.
7-8 (citing Erol i1 25: descriptor is calculated for a subset of pixels of the
image, e.g., for the region comprising the face of a person)), teaches the
disputed limitations of claim 1.
Thus, we are not persuaded that the Examiner erred in finding the
combination of Li, Shiiyama, and Erol teaches the descriptor identifiers of
claims 1, 10, and 20.
A Threshold Number of Descriptor Identifiers
Appellants contend the combination of Li, Shiiyama, and Erol fails to
teach identifying an indexed image that has a threshold number of descriptor
4
Appeal2015-002174
Application 12/610,810
identifiers that match one or more of the first descriptor identifiers, as
required in claim 1. Appellants argue Li identifies matching images by
comparing the distance between feature vectors, and Shiiyama matches
images by computing a degree of similarity of color descriptor values, but
neither reference determines matching images using a threshold number of
descriptor identifiers that match. (Br. 9-10.) Appellants argue claim 20 on
a similar basis as these arguments for claim 1. (Br. 18-19.)
Appellants' contentions do not persuade us of Examiner error in the
rejection of claims 1 and 20, because the claims do not require any specific
threshold number of identifiers to be matched. Therefore, we agree with the
Examiner's finding that the image search methods of Li and Shiiyama,
which both return a set of matches ranked in order of similarity of
descriptors, teaches identifying indexed images that match at least a
threshold number (i.e., greater than zero) of descriptor identifiers. (Ans. 5
and 10 (citing Li ,-r 33; Shiiyama ,-r,-r 125-'26).)
IdentifYing a Plurality of Interest Points
Appellants contend the combination of Li, Shiiyama, and Erol does
not teach "identifying a plurality of interest points in the image, each interest
point comprising one of a point, a region, or an area in the image that is
identified by an operator algorithm that operates to analyze the image to
identify the plurality of interest points" as recited in claim 10, because
dividing an image into blocks as taught by Li and Shiiyama is not using an
operator algorithm to analyze an image to identify interest points. (Br. 13-
14.) We do not find Appellants' contention persuasive of error, because we
agree with the Examiner's finding that the broadest reasonable interpretation
5
Appeal2015-002174
Application 12/610,810
of an "interest point" is any area of the image that is operated on by the
algorithm, which includes the plurality of image blocks of Li and Shiiyama
that are analyzed by their respective algorithms. (Ans. 11-13 (citing Li
if 30; Shiiyama if 80).) (See Morris, supra.)
Appellants further contend the combination of references fails to teach
"selecting a subset of interest points for further processing," as recited in
claim 10. As discussed supra, we agree with the Examiner's finding that
Erol teaches selecting a subset of interest points within the image (Ans. 13-
14 (citing Erol if 25: the subset region comprising the face of a person)), thus
we are not persuaded of Examiner error in the rejection of claim 10.
Paid Search Result
Appellants contend the combination of Li, Shiiyama, Erol, and Valz
does not teach "associating at least one paid search listing with at least one
of the descriptor identifiers in the first set to generate at least one paid search
result" as recited in claim 20, because Valz does not describe identifying a
paid search result based on a descriptor identifier that corresponds with an
image-based point of interest. (Br. 19.) Appellants' challenge to the
references individually is not persuasive of error in the Examiner's rejection.
(See Keller, supra.) We agree with the Examiner's finding that Valz teaches
the concept of generating paid search results using paid search listings (Ans.
24 (citing Valz if 30)), and combining this teaching of Valz with the
descriptor identifier search method of Li, Shiiyama, and Erol teaches the
disputed limitation of claim 20.
6
Appeal2015-002174
Application 12/610,810
Remaining Claims
No separate arguments are presented for dependent claims 2-9, 11-
16, 18, and 19 (see Br. 12 and 17). We therefore sustain their rejections for
the reasons stated with respect to independent claims 1 and 10.
DECISION
We affirm the Examiner's decision to reject claims 1-16 and 18-20
under 35 U.S.C. § 103(a).
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
7