Mehmet Ersin. YumerDownload PDFPatent Trials and Appeals BoardApr 2, 202015059067 - (D) (P.T.A.B. Apr. 2, 2020) Copy Citation 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. 15/059,067 03/02/2016 Mehmet Ersin Yumer P5964-US 5089 108982 7590 04/02/2020 SBMC 116 W. Pacific Avenue Suite 200 Spokane, WA 99201 EXAMINER YANG, YI ART UNIT PAPER NUMBER 2616 NOTIFICATION DATE DELIVERY MODE 04/02/2020 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): docket@sbmc-law.com PTOL-90A (Rev. 04/07) UNITED STATES PATENT AND TRADEMARK OFFICE ____________________ BEFORE THE PATENT TRIAL AND APPEAL BOARD ____________________ Ex parte MEHMETERSIN YUMER ____________________ Appeal 2018-008361 Application No. 15/059,0671 Technology Center 2600 ____________________ Before MARC S. HOFF, JENNIFER L. McKEOWN and STEVEN M. AMUNDSON, Administrative Patent Judges. HOFF, Administrative Patent Judge. DECISION ON APPEAL STATEMENT OF THE CASE Appellant appeals under 35 U.S.C. § 134 from a Final Rejection of claims 1–20. We have jurisdiction under 35 U.S.C. § 6(b). We reverse. Appellant’s invention is a technique and system for three-dimensional facial expression generation. A semantic type (such as smiling, yawning, grinning, angry, and so forth) is detected and used to generate the expression. The semantic type is then used to assign transfer intensities (e.g., 1 We use the word “Appellant” to refer to “applicant” as defined in 37 C.F.R. § 1.42 (2017). Appellant states that the real party in interest is Adobe Systems Incorporated. Appeal Br. 3. Appeal 2018-008361 Application No. 15/059,067 2 weights) to corresponding points of a blend shape and a facial expression source to generate the three-dimensional facial expression. For points exhibiting a greater degree of movement, greater weight is given to the blend shape than the facial expression source. For points exhibiting a lesser degree of movement, greater weight is given to the facial expression source. Spec. ¶¶ 4, 5. Claim 15 is reproduced below: 15. In a digital medium environment to generate a three dimensional facial expression from a blend shape and a facial expression source, a system comprising: a semantic type detection module implemented at least partially in hardware to detect a semantic type defining a facial expression of the blend shape; a transfer intensity assignment module implemented at least partially in hardware to assign transfer intensities to the blend shape and the facial expression source, respectively, for individual points of the three dimensional facial expression, the assignment based on the detected semantic type; and a three dimensional facial expression generation module implemented at least partially in hardware to generate the three dimensional facial expression from the blend shape and the facial expression source based on the assigned transfer intensities. Appeal 2018-008361 Application No. 15/059,067 3 The prior art relied upon by the Examiner as evidence is: Name Reference Date Sasaki US 2006/0188144 A1 Aug. 24, 2006 Lee US 2010/0141679 A1 June 10, 2010 Park US 2010/0259538 A1 Oct. 14, 2010 Bourdev US 2013/0121409A1 May 16, 2013 Yang US 2014/0160123 A1 June 12, 2014 Claims 1, 6, 8–10, 13–15, 17, and 20 stand rejected under 35 U.S.C. § 103 as being unpatentable over Park and Lee. Claims 2, 5, and 16 stand rejected under 35 U.S.C. § 103 as being unpatentable over Park, Lee, and Bourdev. Claims 3 and 4 stand rejected under 35 U.S.C. § 103 as being unpatentable over Park, Lee, Bourdev, and Sasaki. Claims 7, 11, 12, 18, and 19 stand rejected under 35 U.S.C. § 103 as being unpatentable over Park, Lee, and Yang. Throughout this decision, we make reference to the Appeal Brief (“App. Br.,” filed May 04, 2018), the Reply Brief (“Reply Br.,” filed Aug. 14, 2018), and the Examiner’s Answer (“Ans.,” mailed July 26, 2018) for their respective details. ISSUE Does the Examiner articulate reasoning for combining Park and Lee having a rational underpinning, in order to support the Examiner’s conclusion that it would have been obvious to combine the references under § 103? Appeal 2018-008361 Application No. 15/059,067 4 PRINCIPLES OF LAW The Examiner’s articulated reasoning in the rejection must possess a rational underpinning to support the legal conclusion of obviousness. In re Kahn, 441 F.3d 977, 988, 78 USPQ2d 1329, 1336 (Fed. Cir. 2006). ANALYSIS Claims 15, 17, and 20 Representative claim 15 recites, in pertinent part, “a transfer intensity assignment module . . . to assign transfer intensities to the blend shape and the facial expression source, . . . the assignment based on the detected semantic type.” Claim 15 then requires generation of the three-dimensional facial expression “from the blend shape and the facial expression source based on the assigned transfer intensities.” The Examiner finds that Park teaches several claim limitations, but admits that Park does not “disclose assigning transfer intensities to the facial expression source explicitly.” Final Act. 3. The Examiner finds that Lee discloses “assigning transfer intensities to the facial expression source,” citing Lee’s teaching that proximate expression images 4a, 4b, and 4c (facial expression source) are mixed with other standard expression images 4. The Examiner concludes that it would have been obvious to modify Park’s teachings “to assign weight to facial expression source as taught by Lee, to generate realistic user’s own face image more accurately.” Final Act. 3. Appellant argues that Park teaches weighting key models and combining them to provide a 3D facial expression, but does not assign transfer intensities to image frames. Appeal Br. 20–21. The Examiner, in the Answer, maintains the rejection, asserting that the rejection is based on the combination of Park and Lee. The Examiner Appeal 2018-008361 Application No. 15/059,067 5 finds that Park teaches determining “a blending weight for each facial region of each key model by using the image frame having the facial region most similar to a facial region of a key model.” Ans. 8. The Examiner finds that “[k]ey models are blended with image frame for each region in fig. 11 step 1130 and then synthesized together in step 1140.” Ans. 9. We do not agree with the Examiner’s finding that Park teaches “blending” key models and image frames. Figure 11 of Park is reproduced below: Figure 11 of Park is a flowchart illustrating an example of a method for generating facial animation. Park ¶ 39. Appeal 2018-008361 Application No. 15/059,067 6 Representative claim 15 recites a transfer intensity assignment module that “assign[s] transfer intensities to the blend shape and the facial expression source” (emphasis added). Step 1120 of Park determines a blending weight for each facial region of each key model, by using an image frame having a facial region most similar to a facial region of a key model. Park ¶ 93. We agree with Appellant that Park does not determine a blending weight for any image frame. Contrary to the Examiner’s finding, Park does not weight an image frame at step 1110; rather, step 1110 involves selecting one image frame that will be used to generate the blending weights of every facial region of every key model. Reply Br. 3. Step 1130 of Park blends (only) key models on a facial region-by-facial region basis by using the determined blending weights. Park ¶ 95. Lee teaches mixing proximate expression images 4a, 4b, 4c with other standard expression images 4 to transform the reproduced expression pictorial/video images 6a, 6b, 6c. Lee ¶ 61. Assuming arguendo that the Examiner correctly characterizes the proximate expression images of Lee as analogous to the claimed “facial expression source,” Lee then teaches assigning weights to (only) facial expression sources. Neither Park nor Lee teaches or suggests assigning transfer intensities (i.e., weights) to the blend shape and the facial expression source, as claimed. The Examiner has not cited a teaching or suggestion in either reference as motivation for the skilled artisan to use the two types of weights in combination, as the claimed invention requires. The Examiner concludes that modifying Park in view of Lee’s “teaching of assigning weight to facial expression source” would “generate realistic user’s own face image more accurately in real time.” Ans. 11. The Examiner does not support this conclusion by citing to any factual Appeal 2018-008361 Application No. 15/059,067 7 evidence. Park generates a facial animation “which makes facial expressions similar to facial expressions of a person included in the image frames.” Park ¶ 95; Reply Br. 5. We agree with Appellant’s argument that, therefore, assigning weights to portions of that image frame would not improve the accuracy of Park’s facial animation relative to the person’s facial expressions in the image frame, “because some features would be given more preference than other features.” Reply Br. 5. Representative claim 15 further recites generating “the three dimensional facial expression from the blend shape and the facial expression source based on the assigned transfer intensities” (emphasis added). As discussed supra, we do not agree with the Examiner’s proffered motivation to combine Lee and Park. Therefore, we conclude that the Examiner has not established motivation to combine the references to generate such a three- dimensional facial expression based on transfer intensities assigned to both blend shapes and facial expression sources, as is required by this element of claim 15. We conclude that the Examiner’s proposed combination of Park and Lee lacks a rational underpinning to support a conclusion that the invention recited in claim 15, or claims 17 and 20 dependent therefrom, would have been obvious. We do not sustain the Examiner’s § 103 rejection of claims 15, 17, and 20 over Park and Lee. Claims 1, 6, 8–10, and 13–15 Independent claim 1, like claim 15, recites assigning “transfer intensities based on the detected semantic type to the blend shape and the facial expression source,” as well as “generating, by the computing device, the three dimensional facial expression from the blend shape and the facial Appeal 2018-008361 Application No. 15/059,067 8 expression source based on the assigned transfer intensities” (emphasis added). Independent claim 10 also recites “assigning . . . transfer intensities to the points of the blend shape and the facial expression source,” and “generating, by the computing device, the three dimensional facial expression from the blend shape and the facial expression source . . . using the assigned non-linear transfer intensities” (emphasis added). The Examiner relies on the same teachings of Park and Lee, and the same rationale for their combination, in the rejection of claims 1 and 10. As discussed supra, we find error in the Examiner’s conclusion that it would have been obvious to combine Park and Lee to obtain the invention under appeal. Accordingly, we do not sustain the Examiner’s § 103 rejection of independent claim 1 and dependent claims 6, 8, and 9, or the Examiner’s rejection of independent claim 10 and dependent claims 13–15, over the combination of Park and Lee. Claims 2–5, 7, 11, 12, 16, 18, and 19 Dependent claims 2, 5, and 16 stand rejected over Park, Lee, and Bourdev. We have reviewed Bourdev, and we determine that it does not remedy the deficiencies noted supra in the Examiner’s combination of Park and Lee. We do not sustain the § 103 rejection of claims 2, 5, and 16. Dependent claims 3 and 4 stand rejected over Park, Lee, Bourdev, and Sasaki. We determine that Bourdev and Sasaki do not remedy the deficiencies noted supra in the Examiner’s combination of Park and Lee. We do not sustain the § 103 rejection of claims 3 and 4. Dependent claims 7, 11, 12, 18, and 19 stand rejected over Park, Lee, and Yang. We determine that Yang does not remedy the deficiencies noted Appeal 2018-008361 Application No. 15/059,067 9 supra in the Examiner’s combination of Park and Lee. We do not sustain the § 103 rejection of claims 7, 11, 12, 18, and 19. CONCLUSION The Examiner does not articulate reasoning for combining Park and Lee having a rational underpinning, in order to support the Examiner’s conclusion that it would have been obvious to combine the references under § 103. DECISION SUMMARY In summary: Claims Rejected 35 U.S.C. § Reference(s)/ Basis Affirmed Reversed 1, 6, 8–10, 13–15, 17, 20 103 Park, Lee 1, 6, 8–10, 13–15, 17, 20 2, 5, 16 103 Park, Lee, Bourdev 2, 5, 16 3, 4 103 Park, Lee, Bourdev, Sasaki 3, 4 7, 11, 12, 18, 19 103 Park, Lee, Yang 7, 11, 12, 18, 19 Overall Outcome 1–20 ORDER The Examiner’s decision to reject claims 1–20 is reversed. REVERSED Copy with citationCopy as parenthetical citation