Ex Parte Arngren et al

Patent Trial and Appeal BoardOct 28, 2016

13408448 (P.T.A.B. Oct. 28, 2016)

UNITED STA TES p A TENT AND TRADEMARK OFFICE APPLICATION NO. FILING DATE FIRST NAMED INVENTOR 13/408,448 02/29/2012 6449 7590 11/01/2016 ROTHWELL, FIGG, ERNST & MANBECK, P,C 607 14th Street, N.W. SUITE 800 WASHINGTON, DC 20005 Tommy ARNGREN 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. 3602-0393 4764 EXAMINER FRUMKIN, JESSE P ART UNIT PAPER NUMBER 2156 NOTIFICATION DATE DELIVERY MODE 11/01/2016 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): PTO-PAT-Email@rfem.com PTOL-90A (Rev. 04/07) UNITED STATES PATENT AND TRADEMARK OFFICE BEFORE THE PATENT TRIAL AND APPEAL BOARD Ex parte TOMMY ARNGREN, JOAKIM SODERBERG, and MARIKA STALNACKE Appeal2015-006859 1 Application 13/408,448 Technology Center 2100 Before ALLEN R. MacDONALD, JOHN P. PINKERTON, and GARTH D. BAER, Administrative Patent Judges. BAER, Administrative Patent Judge. DECISION ON APPEAL 1 Appellants identify Telefonaktiebolaget LM Ericsson as the real party in interest. Appeal Br. 2. Appeal2015-006859 Application 13/408,448 STATEMENT OF THE CASE This is a decision on appeal, under 35 U.S.C. Â§ 134(a), from the Examiner's Final Rejection of claims 1, 3, and 5-27, which are all the pending claims. 2 Final Act. 4, 23. Claims 2 and 4 have been cancelled. We have jurisdiction under 35 U.S.C. Â§ 6(b ). We reverse. BACKGROUND A. The Invention Appellants' invention is directed to "[a] method, medium, and apparatus ... for indexing multimedia content by a computer." Abstract. Claim 1 is representative and reproduced below, with emphasis added to the disputed element: 1. A method for indexing multimedia content, the method compnsmg: segmenting the multimedia content, by a computer, into a plurality of segments; identifying, by the computer, for each segment, one or more features present in the segment, wherein the features are of respective media types; identifying, by the computer, for each identified feature in each segment, one or more respective keywords associated the identified feature; and determining, by the computer, for each identified keyword associated with an identified feature in a given segment, a respective relevance of the keyword to the given segment, wherein the respective relevance is dependent on a 2 Claim 5 was also objected to because claim 5 is dependent on claim 4, which has been cancelled. This objection is not before us. 2 Appeal2015-006859 Application 13/408,448 weight associated with the respective media type of the identified feature. Appeal Br. 15. (Claims App.). B. The Rejections on Appeal The Examiner rejects claims 1, 3, and 6-27 under 35 U.S.C. Â§ 102(b) as anticipated by Nevenka (US 2003/0108334 Al; publ. June 12, 2003). Final Act. 4. 3 The Examiner rejects claim 5 under 35 U.S.C. Â§ 103(a) as unpatentable over Nevenka, in view of Jin (US 2012/0158713 Al; publ. June 21, 2012). Final Act. 23. ANALYSIS Appellants argue N evenka fails to teach or suggest "wherein the respective relevance is dependent on a weight associated with the respective media type of the identified feature" as recited in claim 1. 4 See Appeal Br. 8. As argued by Appellants, even assuming arguendo that a conditional probability distribution associated with a feature by the Bayesian 3 Although the header of the anticipation rejection indicated that claim 5 was also rejected, the body of the rejection did not reference claim 5. See Final Act. 4--22. Thus, we treat the inclusion of claim 5 in the header of the rejection as a typographical error. 4 Appellants further argue Nevenka fails to teach or suggest "determining by the computer, for each identified keyword associated with an identified feature in a given segment, a respective relevance of the keyword to the given segment," as recited in claim 1. See Appeal Br. 6-8. Appellants also separately dispute the rejection of claims 23 and 24. See Appeal Br. 11-13; see also Reply Br. 3-5. We do not reach these arguments because the identified issue is dispositive of the appeal. 3 Appeal2015-006859 Application 13/408,448 probabilistic analysis teaches the "relevance" of claim 1, Nevenka still fails to teach or suggest that the conditional probability distribution is dependent on a weight associated with a respective media type. See Appeal Br. 8-10 (citing Nevenka i-fi-183, 84, 88, Fig. 3); see also Reply Br. 2-3. Figure 3 ofNevenka is reproduced below. ---- ---,...._ ____ . . Visual,..._- Audio Transcnpt Figure 3 depicts a directed acyclical graph ("DAG") of a Bayesian network that extracts and organizes features of content streams into three consecutive 4 Appeal2015-006859 Application 13/408,448 layers: low-level layer A, mid-level layer B, and high-level layer C. Low- level layer A includes low-level features (such as signal-processing parameters), mid-level layer B includes mid-level features (such as visual, audio, and transcript features), and high-level layer C includes high-level features (such as segment categorizations). See Nevenka i-f 65. Respective nodes within the DAG represent the low-level features, mid-level features, and high-level features, and links describe a direct causal relationship between the nodes, where a strength of the links is defined via conditional probability distributions ("CPDs"). See Nevenka i-f 83. We agree with Appellants that Nevenka fails to teach or suggest "wherein the respective relevance is dependent on a weight associated with the respective media type of the identified feature," as recited in claim 1, and thus, we are persuaded that the Examiner erred. More specifically, we agree with Appellants that N evenka fails to teach or suggest that a CPD assigned to a link that associates a feature and a categorization is dependent on a weight associated with a media type of the feature. See Appeal Br. 9-10 (citing Nevenka i-f 83, Fig. 3); see also Reply Br. 3. We disagree with the Examiner's conclusion that simply because Nevenka teaches a Bayesian probability graph that is graphically directed from a media type (e.g., "Visual") to a feature (e.g., "Face," "Keyframe," or "Videotext"), and subsequently to a categorization (e.g., "Dangerous Event"), Nevenka also teaches the CPD associated with the link between the feature and the categorization is dependent on a weight associated with the media type. See Ans. 7-8 (citing Nevenka, Fig. 3). Thus, we conclude the Examiner failed to show that N evenka teaches or suggests all the claim elements of claim 1. 5 Appeal2015-006859 Application 13/408,448 We further conclude that the Examiner failed to show Jin cures the aforementioned deficiency of N evenka. Accordingly, we do not sustain the Examiner's rejection of independent claim 1. We also do not sustain the Examiner's rejection of independent claims 12 or 18, which recite substantially similar limitations as independent claim 1, or dependent claims 3, 5-11, 13-17, and 19-27. DECISION We reverse the Examiner's rejection of claims 1, 3, and 6-27 under 35 U.S.C. Â§ 102(b). We also reverse the Examiner's rejection of claim 5 under 35 U.S.C. Â§ 103(a). REVERSED 6